Research on Path Planning of Mobile Robots Based on Improved Q-learning Algorithm
Aiming at such problems as slow convergence speed,long running time and poor learning efficiency in the application of traditional Q-learning algorithm in path planning,an improved Q-learning algorithm combining artificial potential field method and traditional Q-learning algorithm is proposed.The gravitational function and repulsion function of the artificial potential field method are introduced by the algorithm,the reward value is dynamically selected by comparing the gravitational function,and Valueλis calculated by comparingthe repulsion function,the value Q is dynamically updated so that the mobile robot can make explorations with purposes,and can preferentially choose the position far away from the obstacle to move.The simulation experiment proves that,compared with the traditional Q-learning algorithm and the introduction of gravitational field algorithm,the improved Q-learning algorithm speeds up the convergence speed,shortens the running time,improves the learning efficiency,reduces the probability of collision with obstacles,and enables the mobile robots to find a collision free path quickly.
mobile robotspath planningimproved Q-learningartificial potential field methodreinforcement learning
国家科技期刊平台
NSTL
万方数据
