A study of emergency collision avoidance by steering angle designed with energy-optimal control theory

Abstract

The emergency collision avoidance and trajectory planning are investigated in this paper. These are considered to be essential technologies for self-driving cars, which are important in regional revitalization. The collision avoidance by braking is called AEB (Autonomous Emergency Braking), whereas collision avoidance by steering is called AES (Autonomous Emergency Steering). The present research focuses on AES and aims to develop effective control theory based on EOC (Energy Optimal Control theory) with trajectory planning. A theoretical design method for EOC's virtual physical quantities and a method for accurately specifying the target trajectory were proposed. An emergency collision avoidance problem was designed to avoid collisions with vehicles crossing the center line and bicycles encroaching on the roadway on a two-lane road, and the effectiveness of proposed control theory and trajectory planning method are examined. It is found that the present theory can track the specified target trajectory very accurately. And the trajectory planning method proposed here is very simple and can define discontinuous trajectory which is critical for the design of complicated collision avoidance trajectory.

Keywords

• optimal control theory
• real time control theory
• self-driving
• autonomous emergency steering
• position control
• path-planning
• regional revitalization
• trajectory planning
• discontinuous trajectory