An Adaptive Fusion Path Tracking Strategy for Autonomous Vehicles Based on Improved ACO Algorithm

Path tracking system is a key component in autonomous vehicles research. It is a challenge for a single controller to achieve accurate tracking in complex scenarios with dynamic curvatures and errors. Although methods based on dynamic models and optimization theory can improve tracking performance, most autonomous systems lack high-fidelity models and the complexity of optimization processes lead to increase computational burden. Hence, the aim of this study is to develop an accurate tracking control strategy for autonomous vehicles in complex scenarios by proposing an adaptive fusion control tracking scheme based on an intelligent optimization algorithm. Firstly, this scheme is built with a PP algorithm with forward-looking distance and a PID model for direct error feedback as the base controller. Secondly, the PID and PP algorithms are integrated through the ant colony optimization (ACO) algorithm, with adaptive fusion to adjust the weights and reduce tracking errors quickly and effectively. Finally, an improved ACO (IMACO) algorithm is designed by establishing the natural logarithm function to address the blind search problem in the ACO algorithm. Experimental results showed that the IMACO-based fusion controller achieves significant enhancements in path-tracking accuracy, with root mean square error (RMSE) reductions of 68.3%, 74.5%, 35.8%, and 21.8% when compared to PP controller, PID controller, original ACO-based fusion controller, and the Cuckoo-based fusion controller, respectively. This study of path tracking strategy for autonomous vehicles can be devoted to providing a new perspective of the controllers’ development to improve computational efficiency and its application in complex traffic scenarios.