Search

Microscopic traffic simulation models require the calibration of car-following (or vehicle-following) models. The parameters of vehicle-following models control individual driver’s spacing, time gap, speed variation and acceleration during different driving conditions. In recent studies, these parameters have been determined for different vehicle classes separately since heavy vehicles generally keep longer spacing and time gap than light vehicles. These parameters have been commonly estimated based on the observed macroscopic traffic flow data such as average volume and speed. However, these data cannot reflect actual vehicle-following behavior of individual vehicles. Also, the effect of the lead vehicle class on the following vehicle’s behavior has been neglected in the parameter estimation. Thus, this study estimates the driving behavior parameters for cars and heavy vehicles in the Wiedemann 99 vehicle-following model. For the estimation, 2169 vehicle trajectories were obtained from a 640-m segment of US-101 in Los Angeles, California during the morning peak hours. Separate parameters were estimated for three vehicle classes (cars, heavy vehicles, and motorcycles) and three vehicle-following cases (car following car, car following heavy vehicle, and heavy vehicle following car). From the comparison of the driving behavior parameters between cars and heavy vehicles, it was found that heavy vehicles keep longer spacing and time gap with the lead vehicle, are less sensitive to the lead car behavior, and apply smaller acceleration when they start from stationary position compared to cars. It was also found that the parameters significantly varied across different vehicle pairs even for the same vehicle class and the same vehicle-following case. The estimated parameters were also validated as the VISSIM simulation with the estimated parameters better reflected the observed cumulative average speed and acceleration distributions than the simulation with the default parameters. The results indicate that differences in the parameters among different vehicle-following cases and the variability of parameters for different vehicle pairs must be considered in the fixed parameters for each vehicle class currently used in the Wiedemann’s model. [ABSTRACT FROM AUTHOR]

Abstract: This paper examines automated control strategies of variable speed limits that aim at reducing crash potential on instrumented freeways. A real-time crash prediction model was developed to estimate crash potential based on short-term variation of traffic flow characteristics. A microscopic traffic simulation model was used to realistically simulate changes in traffic conditions as an effect of variable speed limits and combined with the crash prediction model for the evaluation of control logics. Within this integrated evaluation framework, the study investigated the effect of strategy control factors on the crash potential reduction and total travel time. The study results indicated that variable speed limits could reduce crash potential by 5–17%, by temporarily reducing speed limits during risky traffic conditions when crash potential exceeded the pre-specified threshold. [Copyright &y& Elsevier]