ABSTRACT
This paper describes the development and simulation verification of a direction control algorithm for autonomous vehicles in consideration of mutual accordance in mixed traffic conditions. This algorithm includes a driver model and a vehicle dynamics model. The driver model is composed of a risk potential driver model and a second-order prediction driver model with multiple preview points. We examined the capability of the algorithm to express human operation. In a traffic condition where there are multiple obstacles, we suggested an algorithm for real-time control target generation from the surrounding environment by using the risk potential and confirmed its effectiveness by simulation and experiment.
