Research on Automatic Lane Changing Method for Electric Vehicles Based on Deep Deterministic Policy Gradient Algorithm

Abstract

Intelligent driving is an important feature of electric vehicles, and automatic lane-changing is an important auxiliary operation in the driving process. It involves complex data such as the external environment, vehicle status, and relationship characteristics of other vehicles. This paper proposes research on automatic lane-changing methods for electric vehicles based on a deep deterministic policy gradient algorithm aiming at the complexity of automatic lane-changing for electric vehicles. First, use the actor-critic model of deep reinforcement learning to realize the design of an automatic lane-changing algorithm. On this basis, the actor-critic model is further improved, and a deep deterministic policy gradient algorithm is proposed and applied to the automatic lane-changing strategy of electric vehicles, which improves the accuracy of automatic lane-changing and thus ensures the safety of vehicle driving.