基于改进型Q-Learning算法的路径规划系统研究

Research on Path Planning System Based on Improved Q-Learning Algorithm

随着无人驾驶领域的兴起,人工智能、强化学习等概念开始普及。人工智能设备具有集成度高、可训练性以及可编程性等特点,在无人驾驶中的路径规划领域发挥了重要作用。论文首先介绍了现有研究中较为经典的路径规划算法,并针对Q-Learning算法效率低下等问题进行研究,提出了一种改进型Q-Learning算法。该算法首先对智能体的运动以及空间环境进行建模,其次改进了Q-Learning算法的奖励机制,最后规定了智能体的运动方式。仿真结果表明,基于改进型Q-Learning算法有效改善了智能体的运动路径以及工作效率。

Artificial intelligence and reinforcement learning have become prominent as the field of unmanned driving has grown in popularity.Artificial intelligence equipment has a high level of integration,trainability,and programmability,and it is used extensively in the field of unmanned driving path planning.This paper first reviews previous research on the classical path planning algorithm,then investigates the low efficiency of the Q-Learning method and presents an improved Q-Learning algorithm.The approach models the agent's movement and environment first,then designs the Q-Learning algorithm's reward mechanism,and lastly specifies the agent's action.The simulation results show that the improved Q-Learning algorithm can effectively improve the movement path and work efficiency of the agent.