Path Planning of Self-driving Vehicles Combining Ant Colony and DWA Algorithms in Complex Dense Obstacles

Abstract

INTRODUCTION: To solve the problems of low quality and weak global optimization of the DWA algorithm, especially the problems of unreasonable path planning and the inability to give consideration to speed and driving safety in the process of vehicles passing through dense obstacles, this paper proposed an improved DWA algorithm based on ant colony algorithm.
 OBJECTIVES: The traffic capacity and computing efficiency of Self-driving Vehicles in complex dense obstacles can be greatly improved.
 METHODS: Through the obstacle density and distance information obtained by high-precision sensors on the vehicle, the speed objective function is updating in real time by using ant colony algorithm. And the maneuverability and safety performance of vehicles passing through are considering by the way.
 RESULTS: The experimental results show that this method can obviously improve the vehicle's traveling ability and uneven path planning in the case of dense obstacles, and the number of iterations of the algorithm is reduced by more than 16%.
 CONCLUSION: The improved DWA algorithm integrated with the ant colony algorithm can effectively improve the operating efficiency of the algorithm, reduce the distance the car must go around outside the obstacles, and improve Car driving safety. The effectiveness and universality of the improved DWA algorithm were verified through experiments.