Discretionary lane-changing behavior at downstream queues, case of choice modeling

Discretionary Lane-Changing (DLC) behavior significantly impacts various facets of traffic flow, including capacity, shock waves, and safety. Understanding the determinants behind DLC behavior is crucial. In this study, we investigate the factors influencing DLC in congested traffic using a combination of video recording and survey methods. We employ a binary logit model to examine the impact of a wide range of explanatory variables such as speed, lateral distance, aggressiveness, age, and driving styles on DLC behavior. Our estimation results reveal that socioeconomic, driving style, and road environment variables all play a significant role in lane-changing. Specifically, several parameters exhibit positive associations with the likelihood of lane-changing when drivers encounter downstream queues. These parameters include the speed of the target vehicle, law-evading behavior, disregard for yellow traffic signals at intersections, lateral distance from adjacent cars, and facing at least two accidents. Our empirical findings shed light on the factors affecting DLC and can enhance the performance of traffic flow models.