Lane Change of Vehicles Based on DQN

Vehicle stability and speed are both important research directions for autonomous vehicles. In this paper, we outline how to use the DQN method to manage the lane change of autonomous vehicles, and use the Maxmin Q-learning method to study the vehicle's autonomous driving system. This method improves the speed and stability of the vehicle in automatic driving mode under certain conditions. However, the model we use also has great shortcomings. In the case of complicated links and aggressive modes of other car owners, there are still challenges in vehicle safety.