Browse Publications Technical Papers 2020-01-5032

Longitudinal Behavior Analysis of Drivers in Cut-Out Scenes Based on Natural Driving 2020-01-5032

Cutting out of the lane in the process of following a car is a natural driving scene that often happens, and it is also a typical scene faced by self-driving cars. In this scenario, the driver may either change lanes after the car in front or continue in the original lane. Based on the natural driving data, the longitudinal driver response behavior of the scene cut out of the car in front is analyzed. For the driver's longitudinal response behavior, firstly, the acceleration behavior domain related to TV1 (Target Vehicle I) was divided by support vector machine method, and then the acceleration time was represented by the transverse body overlap rate of the Vehicle and the Vehicle in front at the initial acceleration time, and the acceleration response process was represented by the average acceleration and the maximum acceleration. The influencing factors were analyzed by single factor analysis of variance and Pearson correlation test. The results show that the factors that significantly affect the acceleration time include cutting average lateral speed, vehicle type in front, longitudinal ttc-1, turning signal in front and lighting weather. The factors that significantly affect the average acceleration include vehicle speed, longitudinal TTC-1 and cut average transverse speed. Finally, a regression equation is established to describe the driver's acceleration time and response process by using multiple linear regression method. The research results are helpful for the development of humanized and better than human longitudinal control strategy of self-driving cars, so as to improve human acceptance of self-driving cars.


Subscribers can view annotate, and download all of SAE's content. Learn More »


Members save up to 17% off list price.
Login to see discount.
Special Offer: Download multiple Technical Papers each year? TechSelect is a cost-effective subscription option to select and download 12-100 full-text Technical Papers per year. Find more information here.