Predictive Risk Assessment Using Cooperation Concept for Collision Avoidance of Side Crash in Autonomous Lane Change Systems

Abstract

We present an innovative approach to risk assessment using cooperation concept for collision avoidance of side crash for autonomous lane change system. The cooperation concept is proposed to check an available space in the target lane for lane change in a short period of the future. Before moving to the target lane, the ego vehicle needs to check whether the object vehicle is cooperative or not for the lane change. The object vehicles are indexed as cooperative driver (CD) or non-cooperative driver (NCD) through the conditions based on the relative longitudinal acceleration and the collision free time. If the object vehicle is CD, the ego vehicle moves toward the target lane within the original lane for the preparation of the lane change. The ego vehicle should keep its lateral position within the original lane until the situation for the lane change becomes safe. As soon as the risk assessment becomes safe, the lane change is performed. Therefore, lane change can be started in advance compared to the conventional methods. The proposed method was validated via computational simulation and experimental results.