This PDF file contains the front matter associated with SPIE Proceedings Volume 12303 including the Title Page, Copyright information, Table of Contents, and Committee Page.

With the development of my country's economy and society, in a challenging competitive environment, the traditional business model of small and medium-sized enterprises is facing huge challenges. SY Bottled Water Company is one of the largest drinking water producers in D City, with more than 20 large and medium-sized drinking water sales agencies in the city. In the past, the company relied on the experience of the delivery staff to select the delivery route. With the gradual expansion of the market scale, the original delivery strategy can no longer meet the current delivery needs. Based on the above background, this paper establishes a VRP distribution optimization model for a single center and single vehicle type, and proposes a corresponding genetic algorithm to solve it. Numerical simulation results show that the model proposed in this paper has been significantly improved in terms of the total distance of the distribution route and the total cost, thus verifying the effectiveness of the model.

In the process of developing petroleum, fracturing development of low-permeability reservoir is a key content. In the process of fracturing development of low-permeability reservoir, in order to realize the reasonable formulation of fracturing scheme and ensure the crude oil production and production efficiency of low-permeability reservoir, oil enterprises need to carry out numerical simulation of reservoir to ensure the scientific exploitation of oil. Based on this, this paper analyzes the reservoir numerical simulation in the process of fracturing development of low-permeability reservoir, so as to provide scientific reference for oil production in this kind of reservoir.

Torsion members are very common basic stress members while the current specification for the torsion reinforcement of concrete members is rarely involved. This paper conducts a finite element calculation study on using the carbon-fiber reinforced composite to reinforce concrete torsion column, analyzes the fracture development rules and destruction form of the members, studies the influence of the two important parameters, that is, axial pressure ratio and paste layer on cracking torque and ultimate torque, and summarizes the method of using carbon-fiber reinforced plastics to give antitorsion reinforcement to concrete column, so as to provide reference for torsion reinforcement of concrete structure as well as research data and reference suggestions for the revision of the technical specifications in the future.

Science and technology is the engine of economic growth and the main driving force to improve comprehensive national power, which promoting the transformation of scientific and technological achievements and acceleratingtheindustrialization of scientific and technological achievements and have become the new trend of science and technologypolicies in the world. In order to promote the visualization of the transformation results of scientific and technological achievements, this study designs and develops a GIS system of map service for transformation of scientificandtechnological achievements based on web standards, combined with geographic information system(GIS) development technology, map visualization technology, and data spatial analysis methods. The system includes national regional distribution map, Beijing regional distribution map, core area regional distribution map and sub-center regional distribution map, which can visualize the distribution of science and technology achievements transformation projectsinthe country, Beijing, core area and sub-center. The system aims to reduce the idle rate of scientific and technological achievements and promote technology application, technology promotion and economic development.

With the rapid transformation of computer hardware and algorithms, mobile networking has evolved from low datacarrying capacity and high latency to better-optimized networks, either by enhancing the digital network or using different approaches to reduce network traffic. This paper discusses the big data applications and scheduling in the distributed networking and analyzes the opportunities and challenges of data management systems. The analysis shows that the big data scheduling in the cloud computing environment produces the most efficient way to transfer and synchronize data. Since scheduling problems and cloud models are very complex to analyze in different settings, we set it to the typical software defined networks. The development of cloud management models and coflow scheduling algorithm is proved to be the priority of the digital communications and networks development in the future.

The Cox models have been widely used for analyzing right-censored survival data based on the partial likelihood functions. Despite its success in estimating the regression parameters, the partial likelihood ignores the existence of the unspecified cumulative hazard rate. This article focuses on the complete likelihood function which considers the estimation of both the regression parameters and the nonparametric components. To estimate both parametric and nonparametric components, we propose the minorization-maximization (MM) algorithm to separate the regression parameters and nonparametric components. Then the high-dimensional objective function is decomposed into a sum of low-dimensional functions which avoids the difficulty of large matrix inversion. Simulation studies demonstrate that the algorithms perform well in various cases. The breast cancer data are used to illustrate the proposed algorithms as well.

Abnormal intraday fluctuations in financial markets may be related to major economic events. Therefore, by analyzing the relationship between the anomalies of high-frequency market data and the attributes of market internal events, market participants can have a deeper understanding of the dynamic characteristics of the market. As a new data model, uncertain data is widely used in finance, location-based services, mobile object monitoring, sensor networks and many other types of applications. In fact, cluster analysis is a data processing process, which uses the characteristic attributes between data and adopts some reasonable similarity index to measure, and finally obtains a mathematical classification result. As for the clustering results, we require that the data object gap in the same cluster should be as small as possible, while the data gap in different groups should be as large as possible, while the data gap in different groups be as large as possible. Differences between data objects are usually defined by attributes of data objects, for example, the distance between data can be defined for numeric attributes.

A decision model of maneuver schemes of multiple targets is proposed in this paper, and a multi-objective optimization algorithm is used to solve the model. There are two problems to be solved in this model. One is to quantitatively evaluate the scheme from the ship to the target, and the other is to select the optimal one from the scheme matrix formed by the combination of multiple ships and multiple targets. Therefore, the prediction results of the two types of maneuver schemes approach maneuver and occupy position maneuver are used as the basic scheme unit of the ship to the target in this decision model. In the analysis process, the opinion of experts is used to evaluate the ships, targets, and decision-making objectives which starting from the four factors: distance, time, angle, and the priority of every single scheme. Then, multi-objective optimization is used to solve the scheme matrix formed by the combination of multiple ships and multiple targets. Finally, we find the global optimal scheme combination that meets the experts’ preferences from the scheme matrix.

In order to solve the problem of large mean square error of regression coefficient in traditional Bayesian high-dimensional data analysis, sir model program is applied to realize Bayesian high dimensional data analysis. This paper capture Bayesian high dimensional data, and use sir model program to set super parameters. The prior probability of Bayesian high dimensional data analysis is calculated, and the topological structure of Bayesian high dimensional data analysis is constructed to realize Bayesian high dimensional data analysis. The results show that the mean square error of regression coefficient in the experimental group is significantly lower than that in the control group, which can solve the problem of large mean square error in traditional analysis.

This article analyzes the common technologies of logical data block of big data analysis system, including big data analysis technology, HBase structure, Spark structure, TPCx-BB structure , combined with the application advantages and application points of logical data block affinity scheduling algorithm , by studying the implementation of hot spot determination rules, JLQ scheduling algorithm, LAS scheduling algorithm, etc., the purpose is to give full play to the application advantages of affinity scheduling algorithm and improve the application value of system logic data blocks.

In this study, a water resource allocation method based on configuration relationship matrix and Excel was proposed. First, the relationship between water sources and configuration units was described using a configuration relationship matrix, where matrix X denotes the sequence of water supply to configuration units receiving the water source, matrix Y represents the sequence of water supply to configuration units from the water source, and matrix Z is the water supply coefficient to the configuration units from the water source. Next, the detailed steps and flowcharts of available water supply, water demand calculation of configuration units and reservoir water balance calculation were given. Afterwards, the corresponding worksheets were set in Excel for the sake of data storage, configuration calculation and outcome statistics. Finally, this method was applied to the water resource allocation of the 13th Division of Xinjiang Production and Construction Corps. Characterized by clear logical structure, high intuition and easy operation, the proposed method is applicable to small-scale long-series water resource allocation calculation with simple configuration relationships.

As a new economic form, digital economy has become an indispensable driving force for the transformation and upgrading of traditional industries. With the advent of the digital era, big data automobile marketing is gradually recognized and widely used by major automobile companies. Therefore, in the face of the changing and uncertain sales market, auto sales stores need to speed up their own digital transformation and development, and make full use of big data technology to achieve the purpose of improving store profits and reducing costs and increasing efficiency. It is necessary to improve its own workflow internally, and realize the fine management and the integration of its resources. It is necessary to build smart stores and digital marketing systems externally, so as to gain a higher insight into customer needs and optimize service experience. As far as possible, data collection, integration, processing and transformation can be realized in every value chain and scenario, so that data can flow in this process, and the development trends of data assetization, service business and asset service can be realized, and our own big data smart store and automobile marketing system can be built.

Edge computing has been applied to all aspects of our life, and has a good development prospect. But the problems of “isolated data island” and “data privacy security” hinder the development and application of technology. Therefore, based on federated learning, this paper proposes a privacy-preserving method based on secret sharing and homomorphic encryption, which can enhance the privacy protection of edge computing without affecting the accuracy of the model. At the same time, theoretical analysis proves that this method can resist collusion attacks and reasoning attacks.

Internet of things (IoT) uses perception technology and intelligent equipment to perceive and identify the physical world, interconnect through network transmission, and perform calculation, processing and knowledge mining. Millions of IoT devices are connected to IoT, generating a large amount of data, and the application of data mining technology is gradually deepening. Meanwhile, the privacy protection aspect in IoT environment has become one of the hot issues. In this paper, by describing the characteristics and privacy threats of IoT large-scale data, analyzing the shortcomings of existing data mining privacy protection methods, we propose a data perturbation privacy protection method based on Gaussian mixture model. The method generates and discloses a set of new data with independent and same distribution as the original data to perturb the original data, which not only effectively protects the privacy of the original data, but also maintains the statistical characteristics and has similar accuracy to the mining model generated on the original data.

With the acceleration of modern process and the development of music, the situation of Chinese traditional music culture resources is grim. With the development of history, many excellent traditional music works will gradually disappear.Therefore, it is urgent to use modern high-tech technology to save the traditional music forms, which has a certain digital storage and management significance.This paper analyzes the current situation of modern traditional music, on the basis of the traditional music database system design, has completed the formulation of music data standards, aiming to save the current situation of traditional music culture, so as to continue the excellent Chinese traditional music culture.

With the development of cloud computing and big data related technologies, it becomes more and more important that the system ability of application-oriented talents under the background of new technologies. Based on the cloud platform and the idea of community cooperation, build a special teaching service cloud platform, which can solve complex engineering problems, realize the system design scheme and implement the application system. Combined with the talent thought under OBE mode, an applied computer professional practice system is gradually formed. Through the practice of the environment provided by the cloud platform, the training effect is enhanced and the improvement of system capability is highlighted.

With the rapid development of science and technology and medical informatization, medical data has emerged on a large scale. Although medical data can provide convenient conditions for data analysis, data mining and intelligent diagnosis, the huge medical data set involves a large amount of patient private information, and there is a risk of leakage. Many existing medical institutions use databases or cloud servers to store medical data, which are also vulnerable to attacks and data leakage. The second is that third-party service providers may illegally sell the original client data. Aiming at these problems, this paper proposes a differential privacy scheme based on Laplace and summary matrix, which solves the problems of large statistical information error, low accuracy and high time and space complexity of the algorithm in the current differential privacy model. And the scheme has higher availability and is suitable for data of different scales.

Recently, Microservice orchestration has attracted a great deal of attention in IT community. As a breaking technique to redefine our cloud, Microservice orchestration service, such as Kubernetes [1], As development teams have rapidly embraced the technique, the importance of securing the orchestration service has become the focal point. Hence, security practitioners are facing great challenges from the new eco-system.

In this paper, we propose a distributed security management system, KubeRM, for cloud native environment. One of the greatest challenges we are facing is how to effectively and efficiently synchronize security detecting abilities across all machines in our large Kubernetes cluster. Hence, instead of implementing a centralized management system, our proposal focuses on designing a distributed security management system, which achieves seconds level security enforcement across all machines.

In further, we also demonstrate our prototype system, KubeRM, which can effectively enforce not only the exploits detection on vulnerable K8s servers but also detect abnormal network traffic against the orchestrator. It leverages a rules-based detection strategies to profile all orchestrator behaviors.

In order to classify and identify the countries of radar signals, this paper analyzes the characteristics and current situation of radar signal recognition, adopts the decision tree classification algorithm in machine learning, studies the data preprocessing and the construction, evaluation and optimization of decision tree model, and realizes the program design and practical application of decision tree classification algorithm, which provides effective support for improving the efficiency of radar reconnaissance and early warning. The research shows that the decision tree model can effectively classify and recognize radar signals after training and optimization, with an accuracy of more than 96% and an AUC value of more than 0.98.

Existing key agreement protocol for IPSec, WireGuard, TLS and QUIC needs separate control message exchange. OKEP (Identity-based one-pass key agreement and encryption protocol for Secure IP) is a secure and lightweight protocol in network layer based on identity-based cryptography. It aims to reduce the number of message exchanges in the key agreement part. It realizes 0 RTT (Round Trip Time) key agreement in the first connection by sending parameter along with data in every packet. Using this mechanism, OKEP does not need to synchronize state after releasing the secure connection. In order to avoid transferring large amounts of data for authentication, OKEP uses the characteristic of identity is public key instead of certificate to achieve address validation. Unlike IPSec, WireGuard, TLS and QUIC, OKEP designs one packet format to accomplish key agreement, encryption and authentication rather not many packet formats. The experiment results show that compared to other secure protocols, OKEP can transmit less data to accomplish key agreement and address verification.

Unmanned aerial vehicles (UAVs) have also been investigated to act as flying edge nodes due to their efficient mobility. In this paper, we build a collaborative mobile edge computing (MEC) system with multiple unmanned aerial vehicles and the base station (BS), where the UAVs and the BS are deployed as edge computing servers. In this system, we minimize the system delay by jointly optimizing the deployment of UAVs, task offloading, and resource allocation. To provide users with better service quality, a heuristic algorithm FIDE is proposed to solve this problem. First, we determine the location of UAVs by fuzzy clustering, then we get the optimal resource allocation by the Lagrange-multiplier method, and finally we apply the partitioning idea to group users and use co-evolution to obtain the offloading scheme of users. The simulation results show that our proposed approach outperforms other baseline approaches under the same conditions.

Application tasks are highly concurrent in the national earthquake precursor network, and the data resources are distributive and heterogeneous. These two characteristics make the data are difficult to access uniformly in the earthquake precursor network, which prevents the better data sharing. So, this paper carries out studies on the key technologies of the heterogeneous data-oriented task engine which mainly includes the task scheduling, the task analysis and the task execution. We propose a thread pool-based task scheduling method, a three-layer view-based task analysis method, and a factorybased task execution method. The results of the tests show that the designed task engine has the good concurrent data access ability.

The general rehabilitation training effect can be improved by clustering according to the degree of cognitive impairment and targeted rehabilitation training for the same type of patients. A patient clustering algorithm is proposed based on Kmeans to realize the rapid classification of patients with cognitive impairment. This work carried out a two-dimensional K-means clustering experiment and a multi-dimensional K-means clustering experiment using MOCA diagnosis information of patients obtained from the ADNI database as the actual data set. The modified cosine similarity method was used to examine the similarity between the eigenvalues of users in the same cluster in which way the clustering effect of the traditional K-means algorithm was verified on various data.

In the face of the increasing volume of data stored today, various enterprise organization outsources their databases to cloud server vendors for storage and management in order on database management costs. As enterprise organization encrypts their data for security reasons when uploading, this can create certain problems for the usability of the data; it is necessary to implement a kNN solution that supports secure queries over ciphertext. A number of approaches on its privacy protection have emerged, but there are still issues such as efficiency and security. To address these issues, an efficient and secure k-neighborhood query scheme SEkNN (Secure and efficient kNN) over cloud computing was proposed using garbed circuits combining secret sharing and Yao protocol to construct Euclidean computation as well as a secure ordering algorithm that works completely secure and more efficient. Security analysis was also performed to prove that the scheme was secure under the semi-honest model. The final experiment was compared with a lightweight kNN scheme based on edge computing (EBkNN) and verified that the scheme better meets the needs of practical application.

The energy consumption of sewage treatment(ST) process is relatively large. Reasonable process design can reduce energy consumption in an appropriate amount. Reasonable control technology can make up for the deficiency of process design or reduce the difficulty of process realization. Reasonable control can also reduce energy consumption and improve water quality. In this paper, an ACS for ST process is designed, and an improved particle swarm optimization(PSO) algorithm is introduced into the system, so that the system can predict key variables such as BOD and COD that affect the treatment process during sewage treatment, and compare the improved algorithms such as HPSO and SPSO with Predictive performance of the underlying particle swarm algorithm. In order to test the sewage control effect of the system designed in this paper, five key indicators that can reflect water quality are measured, and the measured values are compared with the national standard limit value. It is found that after using the system for sewage treatment, the discharged pollutants are all In line with national standards.

With the rapid growth of Internet applications, cloud computing systems are hosting an increasing number of long services, such as web search, e-commerce, social networking, and online video. These long services often require predictable response times to provide a high quality user experience. In this paper, we first maintain an adaptive sliding window whose size will be calculated online based on the currently monitored request arrival rate changes, and then plan future short-term resource reservations for long services based on this sliding window that can satisfy the tail-latency SLOs. The third challenge is to eliminate tail-delay SLO violations due to bursty workloads at runtime. Tail-delay SLO violations can still occur because work negatives are all bursty and it is difficult to achieve zero error in predicting workloads. This paper addresses this challenge by designing a runtime monitoring mechanism that dynamically adjusts resource allocation based on the statistical characteristics of the response latency of requests. The main contributions of this paper are as follows: (1) A fine-grained resource scaling engine for containerized long services is proposed to minimize the resource waste due to overcapacity reservation while satisfying the tail-latency SLOs.(2) A hybrid resource reservation scheme is proposed that can handle both periodic and acyclic workloads of long services.(3) An online resource redeployment mechanism is designed to eliminate potential tail latency SLO violations due to workload bursts.(4) Implemented a prototype system for HRSE.

In the cloud computing environment, consider how to achieve efficient task scheduling, improve user satisfaction, and reduce task response time and time span. Consider energy consumption saving factors and users This paper also considers how to reduce the energy consumption in the scheduling process and improve the time and service quality of users. The specific research objectives of this paper are as follows. (1) To design an improved resource scheduling algorithm for efficient resource scheduling process in cloud environment for the purpose of reducing the response time and time span of tasks. (2) Propose an energy-efficient resource scheduling algorithm to reduce energy consumption in the scheduling process of cloud computing. An energy-efficient resource scheduling algorithm is proposed to reduce the energy consumption in the scheduling process and achieve load balancing. First, we surveyed the literature related to the resource scheduling problem in the cloud environment and summarized and analyzed the application scenarios and performance advantages and disadvantages of existing scheduling strategies; based on the research results, we divided into two subtopics: research on efficient resource scheduling algorithms under different scenarios and research on energy-efficient resource scheduling algorithms in the cloud environment. The research focus under different sub-topics is clarified; mathematical modeling is performed and scheduling algorithms are designed according to the research problems under the sub-topics; then the advantages and disadvantages of the algorithms are analyzed theoretically; finally, a simulation environment is constructed, and the performance of the proposed algorithms is verified through simulation and compared with existing cloud scheduling algorithms in the same application scenarios.

Focused on the issue that the feature information extracted in malicious code cannot fully explain the behavior functionof malicious code, and the neural network used cannot extract spatial features and time series features at the same time, amalicious code classification method based on API sequences and text convolutional neural network Text-CNNwasproposed. Firstly, the method used the binary file analysis tool Angr to reversely analyze the malicious code binaryfile, obtained its data structure and control flow information, and automatically generated the control flow graph and functioncall graph. On this basis, an API call sequence extraction algorithm was proposed, which could generate the API call sequences according to the sequence of API functions used by malicious code. Secondly, an API call sequencevectorization model was established by using the word2vec model to vectorize the API call sequence, so that eachAPIfunction could obtain a vector representation of itself. Then, a malicious code API call sequence was transformedintoamalicious code API matrix, which was used as the input of the classification model. Finally, drawing on the idea of text classification, a malicious code classification model MM-Text-CNN was proposed. This model combinedone-dimensional convolution operation and two-dimensional convolution operation. It was not only suitable for input dataofdifferent sizes, but also can simultaneously extract spatial and temporal features of input data. The experimental resultsshowed that the classification model proposed in this paper can complete the malicious code classification task, andtheaccuracy rate could reach 97.83%.

The performance of 3D CAD software is usually evaluated based on a lot of test data. However, the complex uncertainties intrinsic in the test data has not been investigated so far, the influence of which on the reliability of performance evaluation result has not been considered either. To address the problem, this paper analyses the complex uncertainties in test data and proposes a visualized evaluating method for the performance of 3D CAD software based on the characterization of test data by cloud model. First, the initial weight of each performance index is determined by analytic hierarchy process (AHP) and the test data are standardized. Then the standardized test data are input into the backward cloud generator to calculate the digital characteristics of the cloud model for each performance index, based on which the reliability of each performance index is calculated and the corresponding weight is adjusted accordingly. Finally, the overall performance of 3D CAD software is scored by AHP and the reliability of the evaluation result is also rated. An analytic hierarchy diagram is produced for illustrating the whole evaluating process and the cloud map of the overall performance is generated by the forward cloud generator to provide the intuitive and reliable evaluation process and result for users.

Side-channel attacks (SCAs) have become one of the main threats to encryption devices due to their low cost, short time, and strong attack capability. CPU is the core of encryption devices. Thus the resistance to SCAs of the CPU is essential for protecting encrypted information. Based on the register-transfer level (RTL) net-list simulation and CPA method, this paper carries out SCAs on the CPU running the AES-128 algorithm. A protection scheme based on random power consumption disturbance strategy is proposed. It uses register configuration to send random pseudo operation insertion requests to some idle modules inside the CPU at a certain frequency, thereby reducing the correlation between the power consumption and encrypted data. This paper innovatively analyses the effects of the number of randomly flipped modules and the insertion frequency of random pseudo operation on the CPU's resistance to SCAs. According to experimental results, when the CPU is not protected, only 30 power traces are able to reveal the correct key. The required trace number increases when the random power consumption disturbance strategy is applied. The anti-attack performance of the CPU is proportional to the number of randomly flipped idle modules, and approximately inversely proportional to the power consumption signal-to-noise ratio (SNR). Particularly, when all idle modules randomly flip at a frequency of 25%, the CPU’s anti-attack performance has been improved by 3700 times. The proposed protection scheme is simple, easy to implement and highly flexible, by taking the safety performance, power consumption, area and other factors into consideration.

Water scarcity is the most serious problem facing a large part of the world's population, especially in the United States. Guided by goal programming and dynamic programming, this study uses MATLAB to carry out the concrete practice of genetic algorithm based on the conflict between water use and power generation water use (agriculture, industry, residential). This method can also be used to study the water use conflict in the optimization process of water resources optimal allocation in other areas. It is calculated that the ratio of pumping of domestic, industrial and agricultural water is 0.98:0.01:0.01, the pumping volume of Lake Powell and Lake Mead is 68,320,154,372.388 m³ and 5,283,944,815.489 m³ respectively, and the water supply quota of domestic, industrial and agricultural water in five big states is analyzed. Firstly, household water consumption accounts for the majority of the total water consumption, so the government should strongly advocate people to save water. Second, industrial water is much more economical than agricultural water, so the second type should meet industrial water needs. Finally, as rainfall declines in the Southwestern United States, the government should consider developing water-saving agriculture.

Cloud computing has received widespread attention and application in the industry due to its strong computing power and large storage capacity. However, while users enjoy the convenience brought by cloud computing technology, they inevitably face the challenge of data security, and how to ensure the privacy security of data in the cloud environment has become an urgent problem. Therefore, this paper investigates the privacy protection methods of data security based on cloud computing environment. This paper first analyses the architecture of cloud computing and the current threats. Then, it focuses on the principles of AES and SM2 algorithms, and on this basis, a hybrid encryption data security privacy protection scheme is proposed by combining the SM2 algorithm with the AES algorithm. Through experimental tests, the results demonstrate that the scheme improves the computational efficiency of encryption and decryption while effectively protecting the privacy and security of data transmission in the cloud environment.

With the popularization of computer technology and network information technology, cloud computing technology is gradually rising, and the integration of many modern science and technology, and then the traditional translation industry has carried out a new information reform. Machine aided translation system based on cloud computing mode, the intersection of traditional machine aided translation system can greatly shorten the system response speed, improve system efficiency, reduce the system operation cost, and the complex translation needs to adapt to the present society and put forward feasible application pattern, provides a perfect machine translation solutions.

In order to solve the problem of high operating cost of maintenance of enterprise caused by unreasonable number of maintenance personnel, using queuing theory to establish the maintenance queuing model of equipment. The Semi-Markov analysis method is used to calculate the one-step state transition probability of the maintenance system. The indexes such as stationary distribution and average queue length are obtained. The problem of the number of maintenance personnel is transformed into the setting problem of the optimal number of maintenance service stations, and based on the marginal analysis method, the scheme of maintenance personnel allocation is obtained to minimize the total loss cost of the enterprise. Through the collection and analysis of the fault data related to a power plant in the enterprise, the optimal maintenance personnel quantity scheme of the enterprise is obtained, which can ensure the lowest unit time operation cost of the enterprise under this scheme.

The development of earth observation technology and the maturity of data storage technology have prompted the remote sensing industry to enter the "big data" era. PIE-Engine is one of the remote sensing cloud computing platforms of China, which can integrate the main processing work involving remote sensing images into the program, so as to realize the rapid and batch processing of massive remote sensing data. This paper studies the application of PIE-Engine in water ecological environment monitoring. It selects landsat8 data in part of the Taihu Lake region from 2013 to 2020, and filters the data, quantitatively inverts the model, and visualizes the results. The program runs on the platform to study the concentration of turbidity of cyanobacteria blooms in the Taihu Lake. The results indicate that PIE-Engine can screen suitable remote sensing images and rely on established models to perform fast band calculations. Finally, the experimental results are exported according to user requirements. It can be seen that PIE-Engine has broad application prospects in water ecological environment monitoring. In the future, more experiments and research are still needed to continuously improve the accuracy and universality of different water quality parameter inversion models, and expand the platform data source and algorithm, give full play to the advantages of the platform and promote its applications.

Aiming at the problems of randomness and fuzziness in the railway emergency，and multiple scenarios in the evolution of the railway emergency, a multi-attribute group decision-making method of railway emergency is proposed based on prospect theory. Firstly, the uncertainty transformation of linguistic decision-making information is realized using the qualitative and quantitative transformation characteristics of cloud model. Secondly, the cloud distance and the cloud possibility degree are defined according to the concept of interval number, then the cloud prospect decision matrices of each decision-maker are constructed. The attribute weights of each decision-maker are obtained through establishing the attribute weight optimization models based on the principle of maximum prospect value of every alternative. Thirdly, the decision-maker weights are determined based on group’s consistency and the maximum entropy, and the best alternative is obtained by comparing the whole prospect value of alternatives. Finally, the validity and feasibility of the method are verified by an example.

Numerical simulation calculation is an important technical application in reservoir engineering. Technicians can use numerical calculation method to solve the problem of liquid permeability in reservoir engineering by compiling computer calculation program. Under the action of this technology, technicians can simulate the effect of reservoir engineering development and select advantageous schemes according to the actual situation. It provides a basis for enterprises to make scientific decisions, and can evaluate, predict and improve the productivity of reservoir engineering. This paper studies the method of reservoir engineering numerical simulation, explains the steps of technology application, and makes a scientific prediction of the future development trend of this technology, in order to promote the development of domestic reservoir engineering numerical simulation and make more new technologies emerge in this field.

The commissioning of Lu-Gao-Zhao and Kun-Liu-Long DC projects has brought about significant structural changes in the main grid of China Southern Power Grid. In view of the complex AC/DC operation mode and inter-provincial complex transaction mode of China Southern Power Grid, this paper systematically carries out the research and application of interregional and inter-provincial dispatching plan system upgrade. Firstly, the characteristics of dispatching plan of China Southern Power Grid before putting into operation of three-terminal DC project are introduced, and the difficulties and adjustments brought to dispatching plan by putting into operation of three-terminal DC project are analyzed. According to the characteristics of three-terminal DC operation mode, the simplified processing mode of three-terminal DC scheduling plan based on feature disassembly is constructed to meet the needs of scheduling plan preparation under the new situation. Finally auxiliary analysis function of intelligent dispatching decision-making plan is developed and implemented spanned across the province to send electricity and power generation plan arrangement, days before scheduling execution and so on are widely used, fully verified the upgraded system, the feasibility and effectiveness of the scheduling plan, significantly improved the southern power grid system security, stability and economic operation level.

Since the construction of energy internet has been continuously moving forward, various sensors that are involved in the energy system tend to increase dramatically. Massive data obtained from those sensors impose huge pressure on the transmission system and heavy computation load on the centre station, while traditional data processing approaches have suffered from undesirable data quality, data losses and format inconsistency. The introduction of edge computing is to redefine the framework of intelligent perception system by achieving data processing locally, and hence optimize the system through cloud-edge collaboration mechanism. In this paper, the optimized transmission approach for multiparameter intelligent perception has been given, after which the design methodology for its edge computing algorithm has been proposed, including data unification, data identification and distributed swarm intelligence.

With the widespread popularization and rapid development of the Internet, the Internet of Things and intelligent equipment in modern construction site production, there are a large number of production equipment in actual production. Once the equipment fails, the equipment is shut down, and the fault is pending repair. It will cause inestimable economic losses, even threaten the lives of employees in severe cases, and bring serious negative impacts to the society. Preliminary prediction of equipment failures and predictive maintenance of failures are of great significance in production and maintenance. Aiming at the problems of preventive maintenance without specific algorithm support, difficult calculation and high cost in the equipment management of the research object, this paper combines equipment life and equipment maintenance and establishes the corresponding equipment life estimate after optimization through an improved genetic algorithm. The model proposes a low-cost and easy-to-obtain method for data analysis samples, which provides powerful algorithm support for preventive maintenance of equipment.

Face swap is to transfer a face image of one person to another given person. It can generate false but very realistic images and video data. At present, numerous frameworks have been proposed, such as CNN, FCN, GAN. Recently, GAN-based methods are popular with satisfying performance. In this paper, we aim to improve the GAN-based face swap methods under the cycleGAN framework and achieve a more realistic face-swapping result. In this method, cycleGAN is applied to to face swap task, which mainly solves the problems of unavailable paired training images, and our model can be trained without pairing real images. More than 1,000 pictures in total have participated in the model training. The test needs to input a video and output a video of the facial changes in a few minutes. We collect Hillary and Trump videos data on the google website for testing and analysis, obtain a trained model, and accurately change the face of the video. Our method can handle face swap effectively with reasonable speed. The quality and speed of its generation are no less than that of other frameworks like GAN, which is more popular nowadays.

Sentiment analysis of traditional deep learning algorithms does not fully consider text features and input optimization. To solve this problem, this paper proposes a deep learning model, namely CNN+ATT model, which combines the Convolutional Neural Network(CNN) model and the attention mechanism. In this paper, the sentiment analysis of the CHNSENTICORP-HTL-BA-6000 hotel review data set is carried out through the training of the convolutional neural network model. At the same time, combined with the deep learning algorithm of word2vec and Attention mechanism, compared with the CNN model, the method in this paper improves the accuracy and F1 value of the model.

As one of the main means of transportation for citizens in Wuhan, urban rail transit has assumed the dual responsibility of ensuring the travel needs of citizens and blocking the spread of the epidemic in the context of COVID-19. Taking the security check space of Wuhan subway Street entrance station as an example, the paper aims at putting forward the optimization strategy of security space design to solve the obstruction problem caused by the excessive flow of subway stations at present. The paper takes the COVID-19 prevention and control requirements in Wuhan into consideration, uses intelligent technology, combines the construction of social force model to conduct pedestrian simulation, and applies simulation variable analysis. The findings indicate that the optimization strategy of security space design effectively shortens the arrival time and effectively controls the flow of people. It is expected to provide some reference and research basis for the design and optimization of security inspection space of subway transportation system in the future.

CPS was first mentioned by NASA in 1992 and elaborated on this concept at the National Science Foundation (NSF) in2006. Taking the third paradigm of intelligent manufacturing as the prototype and discrete manufacturing unit as an example, this paper expounds the unit level CPS system in detail, and forms the system level CPS architecture and the system level CPS architecture of the system.

In the era of informationization, knowledge acquisition has penetrated into people's daily life. Although traditional knowledge service platforms have a rapid knowledge update speed and cover a wide range of fields, they have problems such as low professional threshold, lack of clear logical knowledge structure and difficulty in knowledge rights maintenance and protection. Although knowledge graph can effectively manage knowledge, it faces challenges when it comes to expanding the knowledge base. Based on this, a knowledge system crowdsourcing expansion method is proposed to enrich and improve the knowledge base. The method is integrated with blockchain technology to ensure the openness, transparency and fairness of the crowdsourcing process through smart contracts. In addition, it also improves the authentication and transaction process of knowledge contribution on the blockchain, which solves the problem of difficulty in maintaining knowledge rights in traditional knowledge service platforms and further protects the legal rights of knowledge contributors.

In the field of security, video image processing technology plays an important role. An efficient security system based on video processing technology can effectively improve the efficiency of public security, transportation, justice, retail and other industries. Nowadays, with the development of information technology, video image processing technology can gradually replace manual labor in the field of security. This paper analyzes the point processing, area processing and frame processing in video image processing technology in detail, and conducts in-depth investigation and research on the application of video image processing technology in the field of security.

Background: With the continuous development of global economic integration, the e-commerce industry in the newerais also developing rapidly. In this process, warehousing and logistics have become the development momentumthat promotes the quality and efficiency of the logistics industry. Methods: In order to explore the good development of thelogistics warehousing system in terms of collaboration and digital informatization, and try to find the connectionbetween the two, the method of comparative analysis is adopted. Results: It was found that AI+logistics is mainlyreflected in the integration and innovation of new-generation information technologies such as artificial intelligenceandthe Internet of Things. This integration and innovation has injected new development impetus into the warehousingandlogistics industry. Closing: This paper demonstrates that this integrated innovation also works well in improvingcost economy and making the user experience smarter.

In order to solve the key problem of hop count detection of digital vernier calipers before leaving the factory, and the traditional manual detection method requires high labor and is inefficient, a digital caliper based on convolutional neural network and OpenCV computer vision library is designed. Identify automatic detection systems. Firstly, according to the spatial position and size distribution characteristics of the digital display screen of the digital vernier caliper, the digital display area image of the caliper is collected by calling a drive-free USB industrial camera in opencv-python, and preprocessed, and then the convolutional neural network is used. The (CNN) model conducts data training on the image samples of the digital display area, and finally the test part conducts a digital recognition detection experiment of 20 frames per second on the digital display area. The experimental results show that the digital vernier caliper automatic identification and detection system based on the convolutional neural network model has a digital recognition accuracy rate of 99.32%, and the accuracy fluctuation is only 0.68%, which better realizes the need to log the digital vernier caliper before leaving the factory. Hop detection is performed for this purpose.

With the outbreak of information technology, software development companies provide personalized and comprehensive software systems for users' needs, providing many conveniences to their lives. As the basis of software development, requirements run through the whole process of software or system development and determine the quality of software projects. Requirements traceability is applicable to a variety of development scenarios and optimizes the development process. Establishing a complete requirements traceability can effectively speed up the development process and reduce development problems, which has important practical significance. Therefore, in order to improve the efficiency and accuracy of requirements traceability and facilitate the display of traceability results, an entry traceability algorithm based on requirement knowledge graph is proposed. The model uses BERT+BiLSTM+CRF to extract entities to form triples, and uses the pointer generation network to supplement the entries lacking titles to form the knowledge graph. Combining the theory in complex networks, the importance of entities is sorted, and the similarity between entities in the document knowledge graph is calculated by restarting random walk algorithm. Extract subgraphs related to the target entry according to the characteristics of the nodes and a breadth-first algorithm. Finally, the graph kernel method in the graph matching algorithm is used to calculate the structural similarity and entity similarity between the related entry and the target entry, and determine the relationship. Through the experimental analysis of real requirements documents, the results show that the method can quickly and accurately trace the requirements of the set of input entries.

The pantograph arc is an important factor affecting the safety of urban rail transit trains. Based on the monitoring video of Zhengzhou subway train pantograph, this paper proposes a pantograph arc detection method using correlation filtering algorithm. A double-filter structure is adopted, and two filters are trained according to different characteristics of the pantograph-catenary arc, and the tracking and arc detection tasks are completed through the oscillation of the filters. Experiments were carried out on Zhengzhou subway surveillance video, the result shows the method achieves the expected correct rate, has low implementation cost and fast detection speed, and can be simply transplanted into practical engineering detection applications.

In order to further explore the application scenarios of blockchain in the military field, starting from the basis of blockchain technology, this paper analyzes the development requirements of military blockchain, and focuses on the application scenarios of blockchain technology in the military field from the four directions of battle command and control system, equipment management, logistics supply chain management and human resource management. On this basis, the challenges and Countermeasures of blockchain in military application are given, which will provide ideas and reference for the implementation of military blockchain application.

Polynomial multiplication is the most compute-intensive in the Lattice-based post-quantum cryptography (PQC). This operation can be sped up using the number theoretic transform (NTT). A lot of research is currently being done on how to speed up the NTT algorithm. In light of these considerations, this study presents an NTT accelerator with a greatly simplified butterfly unit (BFU) architecture, a completely pipelined modular multiplication (MM) unit, and two parallel computing units. FPGA implementation is realized to evaluate this work. Through the experimental results, processing speed can achieve 2× improvements with controllable resource consumption.

Radio frequency identification (RFID) tags are applied to individual police equipment to realize its life cycle information management. However, due to the lack of application test methods, the performance changes of RFID tags in practical applications have not been effectively evaluated. This paper introduces the information management process of individual police equipment based on RFID technology and puts forward the application specifications of RFID tags. Air interface protocol compliance defines data coding of stored information in the RFID tag. Application read-write performance is affected by metal, plastic, and nylon materials, equipment structure size limitation, and installation location selection. Collision arbitration of RFID tags at least reaches the packing equipment number. Moreover, the security function protects the application equipment information for an administrative institution. The main test items of these specifications and their test methods are designed based on the analysis of the equipment standards. Therefore, the test results of 10 RFID tags on different new types of individual police equipment are given through experiments.

A personalized clothing recommendation method based on expert knowledge is proposed to address the lack of comprehensive consideration of users' appearance characteristics and dressing occasions in current garment recommendation methods. An apparel recommendation knowledge base is designed to store complex heterogeneous data information in the apparel domain in a structured manner and establish semantic relationships between user characteristics and apparel characteristics labels. A forward chain knowledge reasoning method based on the certainty factor (CF) model is used to realize the intelligent recommendation of apparel in combination with expert rules. The experimental results show that the F-measure of the proposed recommendation method in this paper is significantly improved and has a high accuracy and practical value compared with the traditional recommendation method based on collaborative filtering.

Under the background of "Internet + tourism", many OTA sites have large numbers of users and have accumulated huge amounts of user review information. In this paper, web crawler is used to grab Chinese comment data on OTA website for pre-processing. BERT pre-training model is used to represent text feature vectors at sentence level. The results are then input into LSTM model to obtain high-level semantic features. Thus, the emotional tendency of the comments can be classified. The comparison of experimental data shows that BERT-LSTM model has higher classification accuracy, recall rate and F1 value. It shows that the effectiveness of text sentiment analysis has been improved. An emotional analysis of these emotional commentary texts is conducive to the monitoring of management departments, providing reference for the government and enterprises to make decisions. And this research has very high application value for promoting the development of tourism.

Vibration energy harvesting is a very promising technology for powering wireless sensor nodes. However, the narrow and fixed frequency bands of common vibration energy harvesters greatly limit their widespread applications. In this paper, a mass with movable gravity center is introduced to the cantilever beam of piezoelectric energy harvester for frequency adjustment, and it is used as a power source of an autonomous wireless accelerometer. The frequency of the energy harvester can be adjusted from 130 Hz to 180 Hz. Its output electrical current is rectified, and the electrical energy is stored in an electrolytic capacitor. An automatic voltage regulator circuit is developed as a DC-DC converter to stabilize the output voltage, as well as automatically control the sleep and active mode cycles of the wireless accelerometer. A commercially available MEMS accelerometer is used in the system. An RF transmitter module with a frequency of 315MHz is used for the transmission of acceleration data. The data is wirelessly received by another RF receiving module and recorded in the computer. All parts of the wireless sensor system, including the automatic voltage regulator circuit, accelerometer, and RF transmitter module, are powered by the vibration energy harvester. When the autonomous wireless accelerometer works under the vibration condition of 0.17g @ 177 Hz, the system sends an acceleration data sequence once every 16 s, and the duty cycle of the wirelessly received sensor signal is 0.75%.

Artificial intelligence is a computer science discipline formed on the basis of information network and data processing technology, which involves various fields such as economy and politics. It is highly intelligent and highly applied. Artificial intelligence is the use of computer technology, network communication and other ways to carry out information processing and intelligent analysis, it has a high degree of information processing capabilities, its social development has played a pivotal and irreplaceable role. With the accelerated pace of development in the network era, the employment problem of college students has attracted more and more attention, and at present, there are phenomena such as oversupply of graduates and waste of talent resources on college campuses in China, which seriously hinders the development of human resource management in colleges and universities. The application of artificial intelligence has made a new change to the employment problem of college students, but there are some shortcomings in the process of intelligent development. Based on artificial intelligence technology, this paper puts forward countermeasure suggestions such as improving vocational skills and enhancing their own quality from the technical level.

The term telemedicine was first used in the 1920s, although used many years ago and has continued to evolve today. Medical diagnoses usually require visual information, but remote display systems have recently become a special place due to the constant unavailability of the treating physician or the remoteness of medical centers and the constant need of some patients for round-the-clock care. In this article, an Arduino-based heart rate information system is designed and implemented. Due to the reasonable price and easy accessibility of the created system in the fraction, it has many applications. The results of the designed system showed the system's capabilities to track and know the person's heart rate.

With the rapid development of the smart home industry, the types and numbers of devices in users' homes have increased significantly, making it more and more cumbersome to manage these devices; in addition, with the increase of access devices, the local area network congestion will become more serious, resulting in The delay is too high, and even the control is invalid. In order to easily manage many devices, and to ensure stable and good communication with each device to ensure normal control and monitoring, MQTT-based data transmission and Ble-Mesh networking technology are used to design and develop an iOS-based intelligent home system. And use TLS one-way authentication to ensure the identity authentication and encrypted transmission of each part of the system. It realizes the flexible management of device classification and grouping, and combined with voice control, management becomes simple and convenient; network congestion is avoided; it can also effectively protect user privacy and smart home security

Dynamic taint analysis is a widely used information flow analysis method. The software structure supporting dynamic taint analysis needs running program to monitor the spread of taint, which will bring a large running time overhead. The architecture based on hardware implementation can shorten the running time overhead of taint analysis through hardware modification, which has great research significance. Firstly, we introduced the research background and significance of hardware. Then, we describe the problems of providing support structure for dynamic taint analysis based on software and how to solve or alleviate these problems in hardware architecture on the basis of the introduction of the specific classification and structure of hardware architecture. Finally, on the basis of comparing the characteristics of hardware structure from different dimensions, the future research direction of hardware structure is prospected.

With the continuous development of the Internet of Things, its technical system is also gradually enriched. As a new generation of highly integrated and comprehensive application of information technology, IoT has the characteristics of strong penetration, large driving effect and comprehensive benefits. The application and development of IoT has given birth to a large number of new technologies, new products, new applications and new models, which promote the upgrading of traditional industries and the transformation of economic development, and will become the growth point of future economic development. This paper analyzes the relationship of IoT big data cloud computing platform, discusses the feasibility and core development direction of IoT from the background and concept of IoT, guides the application of IoT in several industries, and provides intelligent information strategy, which is another driving force for the development of information industry after computer, Internet and mobile communication network.

The arrival of the Internet era has spawned a new travel agency business model, namely online travel agency, e-commerce spawned a new travel agency model, which has brought a huge blow to the operation of traditional travel agencies. In addition, although the rise of online tourism is determined by the digital, information era; But more importantly, compared with the single and fixed products of traditional travel agencies, the emerging tourism products highlighting depth, personality and quality can better meet the imagination and expectation of current consumers for tourism. Faced with the impact of online travel agencies, traditional travel agencies have to keep pace with the development of The Times and make full use of information and digital reform.

According to the main problems of spacecraft embedded software design, the universal real-time architecture of spacecraft embedded software without operating system is proposed, and the data flow, control flow, time-sequence design techniques and methods under this architecture are given. The architecture aims at layering and decoupling, and integrates object-oriented and process-oriented programming ideas with task description word as the starting point. The main technical features are : the task modules run independently according to their respective task description word information, and do not affect each other ; task module controlled by the state machine, step by step execution, efficient operation ; inter-layer communication using queues, messages, notifications and other ways, data flow fluent. Finally, an embedded software design example of the universal real-time architecture is given. Compared with the traditional software development model, it has significant improvements in reducing software complexity, improving development efficiency and saving costs.

Guide sign system, which is the main source of identification information, plays a significant role in transmitting identification information. A campus sign system is a systematic combination of characteristic landscape layouts formed in colleges and universities through ordered and unified design according to certain features and then arranged in various campus spaces for teaching and administrative staff for identification. The campus sign system is closely related to students' study and life and generates important effects in various aspects. In this study, the vision calculation, the size dimensions of signs, and the colors and lighting effect in sign systems, which are included in the theory of ergonomics, were investigated. Then, the principles that the sign design for the campus static guide system should abide by were summarized for reference.

Based on the ground Ethernet network, this paper realizes the on orbit maintenance of spacecraft embedded software in real time through the heaven earth link, and overcomes the problems of large delay, low and asymmetric data transmission efficiency, high error rate of data transmission, frequent packet loss and so on. The engineering application shows that it can meet the high reliability and high efficiency requirements of spacecraft embedded software on orbit maintenance function.

The recent development in information and computer technology has offered diverse and far-reaching possibilities for translators and the language service industry. Post-editing of machine translation is the most common solution at present applied by translation agencies to balance cost, speed and quality to meet the future localization paradigm. With the output quality offered by machine translation improving, machine translation has been widely used. Post-editing has been adopted to revise the output of machine translation to bridge the quality gap between that of professionals within a tight project period. This paper investigates translators’ perceptions of and editing actions on post-editing through quantitative method and concludes that post-editing of machine translation is becoming a widely accepted practice and will play a significant role in translation industry.

This paper takes 265 districts divided by New York City as the study unit and uses one month's taxi operation data of New York City to analyze the taxi travel mode and urban functional pattern of New York residents. In the orbit data analysis, the travel time series, order prices and travel volume of each block are extracted from the orbit data, and the travel volume time series data are clustered using the spatio-temporal clustering method to study the spatio-temporal distribution characteristics of passenger travel. Finally, the POI data of New York City is combined. This paper explores the travel rules of different cluster areas and summarizes the urban functional division of each cluster area. The results show that the time series model of taxi travel has significant differences between weekdays and weekends. In addition, weekdays show two peak travel peaks in the morning and evening, which roughly correspond to the commuting peak, and the number of weekend night trips is greatly increased. Quantitative analysis of the obtained regional clusters shows that the differences among the clusters are significant and conform to the actual urban regional functional structure. The results of this study have some significance for the analysis of taxi travel profile and taxi operation management in New York City

In this paper, JAVA virtual machine is discussed through the description of the relevant content, the common classification of exception handling mechanism, and the design of the test system of exception handling mechanism. The key points of design include application of SOOT tool, selection of test algorithm, introduction of JIMPLE technology, abnormal code injection, reasonable selection of different virtual machines and so on. Through the establishment of reliable experiments, the practical value of the test system is evaluated. The purpose is to improve the timeliness of abnormal problem discovery and control the negative impact within a reasonable range.

In recent years, based on application data stream of rapid development, in addition to simple data stream management system (DSMS), complex event processing system (CEP) has been gaining popularity. In this paper, aiming at the application requirements of complex event query and the characteristics of event flow, combined with the characteristics of the Argus system, a complex event processing system prototype Eagle is built. Eagle uses query plan-based approach to detect complex events, uses event graphs to construct event sequences and uses the equivalent node that executes the filter operator as early as possible and shares the event graph for optimization. The sequence of events is extracted as a filter condition, so that many nodes can be shared, which greatly reduces the number of nodes in the event graph. Our experiments show that the optimization strategy improves the performance of the system to a large extent.

To explore the effect of the root system of the caragana on the resistance of loess slopes to freeze-thaw erosion in a high altitude seasonal permafrost zone.This study combined freeze-thaw cycling and direct shear tests to observe the effect of root systems on the shear strength, internal friction angle and cohesion of root-soil composites.In this study, a direct shear simulation of the root-soil complex was established by combining the numerical simulation technology of PFC3D particle flow, and the fine-scale parameters of the root-soil complex after freeze-thaw erosion were calibrated with reference to the indoor test results. This study compares and analyses the effect of caragana root content on the resistance of loess to freezethaw erosion, using loess without roots as a reference group, and investigates the mechanism of caragana root reinforcement of loess. The results show that the freeze-thaw cycle has a degrading effect on the cohesion and shear strength of loess and a strengthening effect on the angle of internal friction; the presence of caragana roots reduces the degrading effect of freeze-thaw on the cohesion and shear strength of loess and reduces the strengthening effect on the angle of internal friction; there is an optimum root content of caragana roots in improving the shear strength of loess and a limit value in improving the cohesion of loess.

In order to investigate the effects of root content and normal stress of arbor root system on the shear strength of root-soil composite, numerical simulations were carried out for four normal stresses and five root contents of root-soil composite specimens by using the finite element numerical simulation method, and the effects of different normal stresses and root contents on the shear strength were revealed.It was shown that: the shear strength of the root-soil composite increased with the increase of root content and normal stress, respectively; the contribution of the root system to the shear strength of the soil was mainly in cohesion.The increase of normal stress weakened the strengthening effect of root system on soil; the effect of root content on shear strength was more sensitive compared to normal stress.The research results can provide a certain theoretical basis for the arbor root system to consolidate the slope and provide technical support for the subsequent development of practical engineering treatment.

In order to quantitatively describe the development of multiple cracks in concrete during dynamic compression, a meso-scale particle element modeling is established to numerically simulate the SHPB stamping test. The box dimension is used to calculate and simulate the development of cracks in the specimen. The fractal characteristics quantitatively described the cracks in the specimen and the region of interest during dynamic compression. Combined with the internal damage variables, the indexes to quantitatively describe the multi crack propagation and aggregate crushing of concrete under the action of high strain rate are proposed, which provides a method to further quantitatively describe the influence of multi crack propagation on the dynamic improvement mechanism of concrete.

In order to verify the rationality and safety of the through tied-arch bridge with inclined steel box arch ribs, the mechanical properties of the bridge are investigated. By establishing a finite element model, its strength, stiffness, and stability are analyzed, and the safety and reliability of the structure can be verified. Combined with the forward-backward analysis method, the suspender cable force of the model is optimized. Compared with the results of the design cable force and the measured cable force in the bridge completion stage, the consistency of the suspender cable force is commented. The results show that the requirements of the strength, stiffness and stability of the structure are met in the construction stage, which proves the rationality of the structural design and the safety in the construction stage. The suspender cable force after model optimization is in good agreement with the design value and the measured value. It provides a reference for the construction and application of such types of tied-arch bridges.

Fog is a common and hazardous weather phenomenon, so it is an urgent problem to predict the development of foggy weather accurately and timely in daily forecasting. In this paper, the ERA5 reanalysis data is used as the background field, and the initial field of the WRF mesoscale model forecast is combined with different physical parameterisation schemes to carry out numerical simulations for an individual case of foggy weather from 22 to 23 January 2021. ① The WRF model can better simulate the generation and extinction of fog. ② When the relative humidity is greater than 80%, the wind is low, the atmospheric structure is more stable and the subtropical high pressure moves northward, it is easy to produce foggy weather. ③ This fog is due to the transport of warm and humid air from the sea surface to the coastal area of a certain place by air currents, which meets the cold air mass of the mainland at the coastal junction.

The rising of machine learning has enabled computers to do difficult tasks that once and only be don e by human beings. However, many applications haven’t taught the machine about the field of art. To make machine more artistic, this paper designed a model based on GAN that enables the computer to generate 2D Japanese animation avatar. In the paper, GAN is found to be an effective method: the generator and discriminator work together to generate normal but real pictures. The experimental result is that in the 300 generated animations, only 3 images are distorted or unreal, indicating that the success rate is 99% and the proposed model has achieved excellent performance. Furthermore, it can be found that most of the generated pictures are beautiful, meaning that the machine is able to draw not only real, but also beautiful pictures.

With the increasingly prominent application of binocular disparity computing technology in mobile terminals, how to achieve efficient and high-precision binocular disparity computing in mobile terminals with limited computing resources has become an urgent problem to be solved. Aiming at the application requirements of the mobile clients, this paper deeply studies the disparity calculation method based on deep learning, introduces disparity refinement to solve the problem of error accumulation, and implements an adaptive multi-stage binocular disparity calculation model. It can make adaptive matching according to the actual situation of application scenarios and computing resources, and the error can be reduced by 2%.

Aiming at the problems of target loss and poor real-time performance in ship detection at sea, a target detection method based on improved YOLOv4-Tiny network was proposed to realize real-time detection of ship video. Firstly, an output prediction scale of 52×52 is added to the network structure to improve the detection ability of small targets on ships. Secondly, based on the improved model structure and self-established data set, the priori anchor frame was redesigned. Finally, the mish activation function is used to improve the model performance of YOLOv4-Tiny. The experimental results show that the improved YOLOv4-Tiny algorithm not only retains the great advantage of fast detection speed, but also significantly improves the detection accuracy. It can realize real-time high-precision detection of ship video, which is of great significance for maritime safety law enforcement and supervision.

Rainfall and flooding have caused urban watershed disasters to become more and more frequent, and urban economic losses have also increased. In this study, the drainage pipe network data, basic topographic data and hydrological data of the Tangxun Lake Basin were sorted out and corrected, and the basic data of the Tangxun Lake Basin were obtained, and the refined flood model of the Tomson Lake Basin was constructed by using Info-works ICM software, and the model parameter rate and verification were carried out with the measured rainfall runoff, and the flood model with high reliability was obtained. The use of hydrological models to assist in the analysis of drainage and flood control planning is a trend of refined management of rain and flood in the future.

Android malware is transformed into gray image and recognized by deep learning model, so as to realize end-to-end detection. This method has the natural advantage of resisting code obfuscation attack. At present, most of the methods used by researchers for gray image recognition of Android malware are roughly processed by a single convolutional neural network, which does not make good use of this advantage. Therefore, this paper proposes a new Android malware detection method, which is different from the traditional detection method XML file and class.dex file is purified, combined and mapped into a gray image, and then the VGG neural network model is used to detect the gray image. At the same time, we take out the models such as mobilenet and alexnet for comparison. The experimental results show that the purification method has good Android malware detection effect, and the VGG model training results have higher detection accuracy than the other two algorithms, and all evaluation indexes have good performance.

Aiming at the low detection accuracy of existing fire early smoke target detection models, this paper designs a fire early fire smoke detection model based on YOLOv5. First, check the data to understand the fire-prone scenes and divide them into two categories: indoor and outdoor; Second, according to different scenes, considering the influence of environmental factors such as light and scale, manually collect and label the smoke pictures in the early stage of the fire; Third, The Focal Loss loss function is used instead to alleviate the problem of unbalanced classification in the data set. Experiments on the self-made early fire smoke dataset show that the improved network mAP value is 2.3% higher than that of YOLOv5. The experimental results verify the effectiveness of the algorithm.

The scale of traffic is rising as a result of rapid economic development, and there are more and more vehicles on the road, which can easily cause road traffic congestion and even accidents. Accurate traffic flow prediction has been a research focus of intelligent transportation systems. Taking emergency steps for prevention and management ahead of time is critical to traffic safety and transportation efficiency. Existing models and methods lack the ability to model both temporal and spatial correlations, and suffer from poor extraction of temporal and spatial features, which leads to reduced prediction performance. In this paper, a fusion model SECLSTM is proposed to implement traffic flow prediction. The model includes long short-term memory (LSTM) network, convolutional residual network and Squeeze-and-Excitation (SE) module. A convolutional residual network based on the attention mechanism squeeze excitation module is constructed and fused with the LSTM network. Experiments are carried out using real data sets, and the findings reveal that the proposed model outperforms existing methods in terms of accuracy and efficiency.

With the continuous advancement of new urbanization, the degree of urban-rural integration has deepened. Thetransformation of urban and rural spatial structure and functional replacement make the orderly integration of migrant workers into cities become a hot topic. By obtaining the questionnaire and interview data of migrant workers' workandlife in Xi'an, using Markov chain and logistic model analysis method, the dynamic change process and randomlawofmigrant workers' urban integration are analyzed. The study found that: (1) Affected by the soft and hard differencestrength, the migrant workers in the urban space node showed a dynamic process of infiltration, exclusion, and balancedstay. (2) Under the interaction of seepage force and repulsive force, the resultant force and the distance fromthecityshow an exponential change rule. (3) Increasing the spatial network nodes of small and medium-sized cities, adjustingthe industrial structure and building a big data platform are the best choices for migrant workers to integrate into cities.

Aiming at the problem of low recall and precision of threshold-based disk failure prediction methods, this paper studies disk failure prediction through machine learning methods. By labeling the disk examples in the data set and balancing the number of examples with different labels, after selecting the features that have a greater impact on the results, the machine learning method of random forest is used to train the prediction model, and the performance of the model is verified. When the random forest method predicts the failure of the disk, the recall and the precision are both higher than those of the threshold-based disk failure prediction method.

Identifying the authenticity of Chinese paintings is an important part of the field of Chinese painting and calligraphy. The identification of Chinese dialect can determine the value of a work. In the past identification of Chinese paintings, the most traditional method is generally used, which is to rely on the experience of experts for identification. This traditional identification method is too subjective and has low accuracy, and there is no objective and credible scientific evidence for the identification results. In recent years, with the development of computer vision technology and machine learning technology, the use of computer technology to identify Chinese paintings has become the focus of academic research. Computer technology can scientifically quantify the empirical appraisal of experts, and identify Chinese paintings through physical or chemical means. This paper makes a detailed analysis of the computer vision technology for authentic Chinese dialect identification, and proposes the computer technology that should be used for different Chinese paintings.

In order to reveal the spread law of fire smoke in underground cable trench, the simulation study on the spread characteristics of fire smoke in underground cable trench is carried out. Analyze the temperature changes of each wall and cable surface of the cable trench, and the influence of different cable laying depths, cable spacing and arrangement on the fire spread of the underground cable trench. The results show that under the action of buoyancy, the smoke formed by cable combustion spreads towards the ceiling and diffuses horizontally with the increasing fire. The temperature above the fire source is affected by the fire, and the temperature is the highest. The farther away from the fire source, the lower the temperature, and the heating speed of the lower wall is faster than that of the sidewall. When the depth of the underground cable trench is 1.2m, the cables are laid in parallel and the cable spacing is 0.2m, it has little impact on the fire spread situation of underground cable trench.

Multi-UAV intelligent system can reduce the casualty rate and improve work efficiency in special tasks. Therefore, its related fields have attracted extensive attention and research worldwide. In this paper, we proposed a Task-based LeaderSelection Algorithm (TLSA) and Topology Maintenance Algorithm (TMA) based on leader-follower model. TLSA supports to establish communication network belonging to each task rapidly through exchange messages between neighboring UAV. And the TMA is used to maintain the network topology during flight. The leaders and followers obtain their own flight coordinates and relevant parameters through publish-subscribe communication. Experiments verify the effectiveness of TLSA and TMA.

Competitive cycling is one of the most popular sports in the world. In order to study the correlation between the drag and dynamic pressure field distribution of competitive cyclists at different velocity, the SST k-ω model was used for CFD numerical simulation in this work. The dynamic pressure field around the cyclist's system is analyzed, and the key monitoring points are set up and the dynamic pressure change trend at different velocity and the change trend of the steadystate dynamic pressure difference of the monitoring points are calculated. The results show that the variation trend of the dynamic pressure difference at different velocity is highly consistent with the variation trend of the drag coefficient, indicating that the dynamic pressure difference is an important source of drag.

Counter-fact is the highest form of causal reasoning. Focusing on counter-fact will provide new ideas for higher order causal reasoning. For counterfactual text detection, two detection methods are proposed. One is the automatic detection method based on classification model, that is, counterfactual detection is carried out on the given annotated text data. In this way, the data imbalance problem is solved first, then word vector is generated by doc2vec, and finally the binary SVM training model is adopted.The second is automatic detection method based on sequence annotation model, namely the counterfactual statements of reason and result orientation, to this sequence labeling method is adopted, the individual words in the text annotation, after processing the data to be included in the BiLSTM after training, and through a layer of conditions with the airport, the resulting text labeling information. Experiments show that the proposed method can accurately detect and locate the cause and result in the counterfactual text.

In the airport fire truck driving training simulator, aiming at the prediction of fire extinguishing agent trajectory of airport fire truck, a fire fighting operation visual simulation model is proposed. Firstly, the real-time simulation model of airport fire truck fire pump system based on neural network is established. After simplified design, the real-time performance of the system is guaranteed. Secondly, particle system and texture mapping technology are used to simulate fire extinguishing agent. Finally, the MFC application is used to establish the view simulation model of the airport fire truck driving training simulator by combining the fire fighting operation of the fire truck, and the data refreshing of the fire pump under different speed is realized through the human-computer interaction interface. Experimental simulation shows that the model has the characteristics of strong real-time and high fidelity, and meets the requirements of airport fire truck driving training simulator.

The biological functions of organisms are realized by protein complexes as carriers, so the identification of protein complexes is of great significance for understanding cell organization and protein function. Most of the existing computational methods mine dense subgraphs as protein complexes on the protein interaction network, ignoring the relationship between the characteristics of individual proteins and the overall protein complex. A Core-Attachment structure on Weighted Dynamic PIN (CAWD) method is proposed to solve these problems. The method identifies protein complexes based on the core-attachment structure by defining the fixed structure of the complex core on a dynamic proteinprotein interaction network, weighted using GO annotations. Experimental results show that CAWD outperforms the compared identification methods on the performance detection metrics of precision、recall、F1 and Acc. Compared with the COACH method, the four performance indicators are improved by an average of 12.53% on the standard protein complex dataset NewMIPS, and by an average of 13.03% on CYC2008.

In view of that people need to spend a lot of time in studying garbage classification knowledge, garbage classification system based on cloud development had been designed, and the system includes functions such as text search, photo identification, common sense of garbage classification and green news, which will help people improve the cognition of garbage classification, and promote work development of garbage classification.

With the continuous development of computer technology, the application of network information technology in physical education and training, and truly realize the excess of information technology system and disclosure mode. At present, with the continuous progress of human society, China's economy is also showing a development trend under this background, and information technology has also made continuous achievements under the economic development. In order to improve the quality and level of the sports training system, it is necessary to fully integrate computer technology to build a more reasonable and effective data analysis mode. This paper analyzes the application of network information technology in the physical education training system, analyzes the modern physical education training, and puts forward their own opinions, aiming to improve the progress of physical education training in China, accelerate the development of sports field, so as to promote the development of sports in China.

With the continuous development of the Internet, information technology has penetrated into every aspect of our life and work, and it is imperative to promote informationization in the supervision industry. However, at present, China's engineering supervision projects still have problems such as lagging information technology, low transparency of work, untimely filing of archives and low supervision efficiency. Therefore, based on the fast-developing computer application technology and Internet of Things technology, this paper designs and develops the engineering supervision management information system in order to track and monitor the site situation in real time through quality control, schedule control, investment control, contract management, safety management, information management and organization coordination, which is conducive to ensuring the smooth development of engineering projects, promoting the information development of the whole supervision industry and promoting the sustainable development of the construction industry.

Along with the new consumption form and the advanced consumption concept is generally accepted, the loan becomes the new consumption means under the new consumption form. The loan business not only brings profits to the institution but also makes the institution face a greater risk of default. Although there are relatively mature forecasting methods in the field of credit default forecasting, credit data is characterized by complex feature relationships and a large number of missing values, and credit default forecasting still faces certain challenges. In this paper, the random forest and XGBoost models are used to evaluate the default risk of personal financial loans. The results show that these two models have high AUC values, so they have good learning and prediction capabilities. In addition, comparing the two, the predictive ability of the XGBoost model is better than that of the random forest.

With the development and progress of China's aviation field, the flight system has become more and more complex, and the requirements for the flight control system are becoming more and more strict. In order to ensure the safety of uav flight experiment and improve the reliability of uav flight control flight system design, China has independently developed a uav flight control semi-physics simulation system that combines uav flight control semi-physics simulation. Based on the independent research of UAV flight control semi-physical simulation system, according on the current situation of UAV system development, it aims to help the development of UAV flight control system, so as to accelerate the design process and progress of UAV flight control system.

With the development of aviation industry, the aviation industry system needs more and more high bandwidth, high reliability, real-time performance and high fault tolerance. Under this background, TTE came into being. This paper studies the time-triggered Ethernet SAE AS6802 protocol, and the synchronization design method of SM, SC and CM is designed. The simulation results show that the synchronization design meets the clock synchronization accuracy requirements of time-triggered Ethernet.

In the field of high-performance computing, some application scenarios make extensive use of bit manipulation. RISC-V foundation issues B extension to reduce the number of instructions during the static compilation. But this extension does not cover the more frequent operations fully. Instruction fusion technology can fuse multiple instructions into one instruction at the decode stage to dynamically reduce the number of instructions. Therefore, this technology can supplement RISC-V B extension uncovered instruction, enhancing the processor’s capabilities for bit manipulation. This paper implements the RISC-V B extension 1.0 version on the Xiangshan processor. For accelerating more bit manipulation, we profiled general benchmarks by the simulator, proposed ten new instructions fusion targets, and implemented a fusion decode unit. Finally, we use Coremark and SPEC CPU 2006 integer to evaluate the performance. The Coremark results show that dynamic instructions reduce by 18.14%, and the performance improves by 12.09%.

More and more researchers begin to study PIM architecture. Many researchers use HBM to research PIM architecture. However, There is no PIM simulator based on HBM for researchers to use. Researchers need to spend much time to develop PIM simulation tools by themselves. At the same time, some PIM architecture based on HBM do not consider whether the designed HBM structure can be adopted by the industry. To solve these problems, in this paper, we develop PIM-HBMSim simulator, which supports HBM1, HBM2 and HBM3 standards. In addition, PIM-HBMSim models the latest PIM-HBM memory which has been in used in the industry, so that the PIM research based on PIM-HBMSim can be more practical and adopted by the industry.

In the design of modern high-performance processors, branch prediction is a very important component, and good branch prediction can greatly improve the performance of the processor. And Branch Target Buffer (BTB) is an important part of branch prediction. BTB is used to store information about branch instructions. Different from storing each branch as a unit, Glenn et al. proposed a BTB that stores branch information in a unit of fetch block, which they call Fetch Target Buffer (FTB). This design can optimize the timing by limiting the branch prediction width. But they did not give specific implementation details. In this paper, we propose a specific FTB implementation, and it is used in the second version architecture of XiangShan high-performance processor. First, we give the fields in the FTB entry in this design and their corresponding functions. Most BTB entries need to store the following information: branch tag, branch type, target address. In addition to these fields in this paper, FTB will also store other information for managing entries and optimizing branch prediction performance. Secondly, we proposed the management strategy of FTB entries, when to add, modify, and replace. After all the instructions in a fetch block are committed, we need to decide whether this fetch block needs to be written to FTB, and which way to write, how to avoid the problem of multiple hits. Under the 14nm process, compared to the 1.8GHz of the first version of the architecture, the frequency of the new architecture using FTB can be increased to 2GHz.

Sharing of scattered, heterogeneous and massive design resources by using the design resources integration and sharing platform can improve the collaborative design efficiency of complex product development process. To make the platform high-performance and extensible, following the principle of "thick cloud and thin end", a high-performance and layered infrastructure of the platform is designed based on emerging information technologies such as microservice, container cloud and load balancing technology, and an extensible application architecture of the platform is designed based on the business application standard system. Based on the above-mentioned infrastructure and application architecture, combined with the business needs of group enterprises, the overall logical architecture and functional architecture of the platform are proposed, which provides all-round architecture support for multi-business scenarios where the platform is suitable for the integration and sharing of design resources of group enterprises.

Microgrid makes full use of distributed energy, reduces energy consumption, and effectively exerts the advantages of distributed generation. Aiming at the problem that the interconnected converter is easily affected by the intermittence and disturbance of new energy sources when operating, the internal electronic devices of the system have small damping inertia and weak immunity. A control strategy with a "power-voltage-current" three-loop structure is designed for the interconnected converter. The active power reference value is generated by the bidirectional droop control, and the damping and inertia of the active power are improved by using the VSG characteristics. At the same time, the small-signal modeling and stability analysis of the "active power-frequency" control based on VSG characteristics is carried out. Through the simulation, the power mutual support between sub-microgrids and the power dynamic balance of the whole system are realized to ensure the stable and reliable operation of the system. Then the feasibility and practicability of the control strategy are verified, which provides a theoretical basis for future research on the coordinated control strategy of the microgrid.

The popularity of electric vehicles has brought great challenges to the deployment of urban charging equipment. Whether the construction process of charging equipment can meet the charging needs of new energy vehicles in cities affects the attitude of potential customers towards new energy vehicles and is directly related to the advancement of the national carbon neutrality process. Aiming at this problem, this paper establishes a new energy charging equipment prediction model based on BP neural network optimized by genetic algorithm. The model predicts the number of urban charging equipment matching electric vehicles based on historical data. The results show that the BP neural network optimized by the genetic algorithm can achieve convergence faster than the single BP neural network, and the error is smaller, which proves the feasibility of the model.

The integrated energy system can break the barriers between energy subsystems. It can also realize the complementarity and collaborative optimization of multiple energy sources, and promote the consumption of clean energy and reduce environmental pollution. In order to improve the capacity of clean energy consumption and reduce carbon emissions and system operating costs, this paper establishes an optimal dispatching model of integrated energy system based on electricity trading, considering the objective function of minimizing the operating cost of integrated energy system. Compared with the traditional optimal dispatch model, the example analysis results of the integrated energy system optimal dispatch model based on electricity trading show that the wind power consumption capacity is increased by 25. 46%. The photovoltaic consumption capacity is increased by 30. 28%, the system carbon emissions are reduced by 177. 65 t, and the operation cost is reduced by 29, 539. 4 yuan. It is of great significance to improve the consumption of clean energy, reduce the operating cost and carbon dioxide emissions of the system by modeling and optimizing the scheduling analysis of the integrated energy system, improve the flexibility and peak shaving capacity of the system, and realize the optimization of unit output and load curve.

The application of Vegetation Growth Simulation Model in water resources field has attracted more and more interest among researchers. This paper introduces a water-limited vegetation growth simulation model, which is built based on the physiological principle of vegetation and Water Resources Allocation and Cycle Model (WACM), a generalized water resources allocation model. The model consists of a developmental stage module, a carbon cycle module and a water cycle module, is a common model applicable to various crops. This paper verifies the lengths of plants during growing period, the mass of dry matter, and the variation range of crop yields by referring literature. The model simulates the mass of dry matter and the yields of some crops under water-limited condition in Tuhai and Majia River Basin. The simulation shows an ideal result close to reality.

The desilting tunnel in the estuary of the Taohe River of Liujiaxia Reservoir can discharge the sediment of the Taohe River directly to the lower reaches of the reservoir, which is the fundamental way to solve the sediment problem of Liujiaxia Reservoir. The desilting tunnel has been built for more than five years, and a large number of measured data of desilting applications have been accumulated, which has created conditions for evaluating the desilting effect and the impact of the desilting tunnel. In this study, a one-dimensional numerical model of runoff and sediment is established to simulate the sediment discharge scenario of the desilting tunnel in the estuary of the Taohe River in Liujiaxia Reservoir and then analyze the effect. The study shows that the desilting tunnel in the estuary of the Taohe River can discharge a large amount of sediment from June to August every year, and that the sediment discharge ratio is positively correlated with the sediment concentration of sediment-laden flow in the Taohe River. Apart from proposing suggestions on the operation of the desilting tunnel, the study will also evaluate the effect of the desilting tunnel on alleviating the sedimentation in front of the dam, reducing the elevation of the sand ridge and the amount of sediment passing through the machine in the future, which can provide support for the safe and efficient operation of Liujiaxia Reservoir.

In order to improve the speed and efficiency of the Department of epidemic prevention and control, this paper uses ARIMA model to train and fit the number of confirmed cases on the basis of the historical epidemic diagnosis information of Guangdong Province. By dealing with the stability of time series, determining the parameters of ARIMA model and testing residual white noise, the ARIMA model is established to predict the number of confirmed epidemic cases, and the number of confirmed epidemic cases in March may 2021 in Guangdong Province is accurately predicted, so as to help the epidemic prevention and control departments improve the accuracy and effectiveness of epidemic control.