A review of essential technologies for collision avoidance assistance systems

Abstract

Compared with automatic collision avoidance systems, collision avoidance assistance systems have attracted more research interest because they help avoid collisions in near-accident situations. However, ensuring the robustness and reliability of collision avoidance assistances is difficult because of problems in reliable environment recognition, accurate collision avoidance decision making, and driver acceptance. This article reviews and analyzes three essential technologies for collision avoidance assistance systems in intelligent environment-friendly vehicles. The architecture of the human-in-loop system is summarized, and available environment recognition methods used in collision avoidance assistances are compared. Moreover, the characteristics of path planning and collision avoidance decision making applied in emergency collision avoidance assistance systems are reviewed, and the advantages of applying haptic share control and yaw moment active intervention during path tracking are analyzed. The emergency collision avoidance assistance systems based on driver behavior models and HSC were found to be the most suitable for high-speed emergency conditions. Such systems are not only highly robust but also more likely to be accepted by drivers.