Investigating the Effects of Electric Bicycle Risk Warning Systems on Motor Vehicle Driver Behavior

To investigate the impact of e-bike risk warnings on motor vehicle drivers’ driving behavior and thereby enhance active safety levels in mixed traffic environments. Based on 2,640 traffic accident cases, five typical risk scenarios involving electric bicycles are constructed (intersection running violations, risky left-turn maneuvers, wrong-way riding, red-light running, and evasive maneuvers). Using sample entropy methods to quantify the indicator data and combining them with generalized estimating equation (GEE) models, we focused on analyzing the factors influencing motor vehicle drivers’ safe driving under information warning conditions. The results indicate that electric bicycle risk warnings can effectively improve the safety and stability of drivers in typical risk scenarios involving e-bikes. Specifically, the average speed safety entropy, lateral deviation distance safety entropy, and acceleration safety entropy decreased by approximately 11.59%, 6.95% and 3.73%, respectively, under warning conditions compared to non-warning conditions. In terms of speed control, driver gender, the clarity and comprehensibility of warning prompts, as well as the interaction effects between gender and both the clarity and comprehensibility of warning information, exert significant influences on speed regulation performance. Regarding lateral stability, driving experience, driver reaction, driving state, gender, and the understanding of warning prompts all have significant impacts on lateral deviation stability. Acceleration control, driving experience, driver gender, and the clarity and comprehensibility of warning prompts significantly affect acceleration behavior. The research findings provide a scientific basis for developing targeted in-vehicle safety warning strategies in vehicle-road coordination environments.