关键词: 无人驾驶, 动态避障, 深度确定性策略梯度, 轨迹跟踪, 梯度优化

Abstract:

In the process of tracking and obstacle avoidance control of driverless vehicles, the controlled object has nonlinear characteristics and variable control parameters. The linear model and the fixed mathematical model of driverless vehicles are difficult to ensure the safety and stability of the vehicle in complex environments, and the driverless discrete control process increases the difficulty of control. To address such problems, in order to improve the accuracy of real-time control tracking trajectory of driverless vehicles, and at the same time reduce the difficulty of the whole control process, the paper proposed a Monte Carlo-depth deterministic policy gradient-based obstacle avoidance tracking control algorithm for driverless vehicles. The algorithm builds a control system model based on a deep reinforcement learning network, and adopts excellent training samples in the strategy learning sampling process. It optimizes the network training gradient with the Monte Carlo method, and makes a distinction between good and bad training samples for the algorithm. The excellent samples are used to find the optimal network parameters through a gradient algorithm, so as to enhance the learning ability of the network algorithm and realize a better and continuous control of the driverless vehicle. Simulation experiments of the control method were carried out in the computer simulation environment TORCS. The results show that the proposed improved DDPG algorithm can be applied to effectively achieve the obstacle avoidance tracking control of the driverless vehicle, and the tracking accuracy and obstacle avoidance effect of the unmanned car under its control is better than that of the deep Q network algorithm and the DDPG algorithm.

Key words: self-driving, dynamic obstacle avoidance, depth deterministic policy gradient, trajectory tracking, gradient optimization