Emergency Steering Collision Avoidance and Path Tracking System for City Bus Considering Yaw Response

Abstract

As the autonomous driving expands to complex road environment, the importance of collision avoidance is emerging across industries and academia, and its application for passenger cars has been actively studied. Because the dynamic characteristics of low-floor buses are different from those of passenger cars, it is necessary to develop a corresponding collision avoidance and path tracking algorithms. In this paper, a novel Pure-Pursuit controller and collision avoidance trajectory planner considering vehicle stability for buses are proposed. First, in order to compensate for the shortcomings of the geometry-based Pure-Pursuit controller, yaw rate gain and yaw rate response analyzed from the actual vehicle data are considered to improve the stability of the bus. In addition, the Conditional Integration-Proportion Integral (CI-PI) controller is designed to reduce the effect of disturbances. Secondly, the trajectory planner is developed taking into account not only the surrounding object information, but also the dynamic limitation of the bus. A group of trajectory candidates including dynamic relation of the lateral motion is generated that satisfies path stability with pole-zero analysis. Then, an optimal trajectory is selected through the cost function which reflects the dynamic constraints. Finally, stability and performance of the proposed controller are verified through simulations and field tests.