自动驾驶水平对驾驶行为稳定时间的影响分析

Despite an absence of pertinent studies, the stabilization of drivers’ states following crucial incidents at varied automation degrees has a significant impact on traffic safety. In order to investigate the effects of automation levels on the stabilization time of drivers’ workload and driving performance, a method for acquiring signal stabilization time was proposed based on the confidence interval of sample mean value. A driving simulation was carried out that featured two crucial events (lead vehicle deceleration and lead vehicle lane changing) and three automation levels manual driving (MD), adaptive cruise control (ACC), and highly automated (HAD) driving. The effective data was collected from 53 participants. The results showed that compared to MD, heart rate stabilization time after the two critical events significantly increased 4.56 s. Moreover, automation levels did not show significant differences in speed, resultant acceleration and standard deviation of lane position of the vehicle after the lead vehicle deceleration. Following the lead vehicle’s lane changing event, the driving performance stabilization time increased by an average of 3.2 s and 5 s in the ACC and HAD conditions when compared to MD. The results provide a theoretical basis for the design of mode and time during the control transition of automated driving.