Vehicle driving behavior modeling based on an improved social force model

Traditional vehicle driving behavior models primarily focus on describing vehicle motion patterns; however, they often lack sufficient explanatory ability regarding the intrinsic mechanisms of driving behaviors, particularly concerning the effective modeling of abnormal driving maneuvers. Therefore, this study proposes a vehicle driving behavior model based on an improved Social Force Model (VSFM). By integrating the combined effects of driving goals, driving states, and driving forces, the proposed model simulates the driver's dynamic decision-making process and interprets both normal and abnormal driving behaviors. The model initially characterizes normal driving behaviors across free driving, car-following, and lane-changing scenarios. Subsequently, it simulates abnormal driving behaviors by incorporating cognitive errors and distracted driving influence factors. Validated through a combination of real-world traffic data and simulation environments, the experimental results demonstrate that the proposed model, based on the improved Social Force Model, not only accurately describes the dynamic characteristics of normal driving but also effectively captures the evolutionary process of abnormal behaviors. Consequently, it provides a modeling approach with significant practical implications for the analysis of driving behavior mechanisms and traffic safety research.