|
Rain-fog environments have an adverse effect on the security and reliability of transportation networks. Microscopic carfollowing models, a prevalent tool, are employed to quantify this negative influence. However, there exists no clear consensus in the literature regarding the appropriate calibration of these models in rain-fog environments. This study gathered traffic flow data in real-world conditions of normal, light rain, heavy fog, and rain-fog environments, and then calibrated the parameters of the GHR, FVD, and IDM models through the utilization of genetic algorithms. Lastly, the calibrated models were then validated through cross-validation. The results indicate that the calibrated GHR, FVD, and IDM models are capable of describing the variability of car-following behavior during adverse weather conditions. The validation analysis revealed that the IDM model had a validation error between 0.39 and 0.45 and more accurately replicated drivers' car-following behavior under rain-fog conditions compared to the GHR and FVD models. This study highlights the necessity of integrating rain-fog factors into existing microsimulation modeling practices, which can serve as a foundation for traffic organization and management in rain-fog environments.
|
