Impacts of visuo-spatial working memory on the dynamic performance and safety of car-following behavior

There are significant differences in the processing of perceived visuo-spatial information by drivers, which reflects their velocity optimization and collision avoidance abilities in car-following behavior. In addition, visuo-spatial working memory provides the driver with the functions of spatial information storage and attention allocation, which is critical in driving decision-making. Therefore, we try to use the data of car-following experiments in real traffic scenes to study the association between driving behavior and driver's visuo-spatial working memory and establish a new visual angle model based on data correlation analysis and dynamic characteristics. The effects of newly introduced attention allocation, visuo-spatial working memory duration, memory sensitivity to visuo-spatial rate of change, and vehicle size factors on the nonlinear and linear stability of traffic flow were analyzed. The above conclusions are verified by small disturbance simulation experiments. Finally, we propose a new collision assessment indicator incorporating visuo-spatial working memory and use it to conduct simulation experiments and statistical analysis on the improvement of the new model in driving safety.