动态环境下基于忆阻强化学习的移动机器人路径规划

Path Planning of Mobile Robots Based on Memristor Reinforcement Learning in Dynamic Environment

为解决动态环境下的移动机器人路径规划问题,提出基于改进蚁群算法和基于忆阻器阵列的DQN(deep q-network)算法的双层路径规划算法。通过改进了概率转移函数和信息素更新原则的蚁群算法完成静态全局路径规划;利用忆阻器“存算一体”的特性,将其作为神经网络的突触结构,改进了传统DQN算法结构,完成移动机器人的局部动态避障。根据移动机器人感知范围内是否有动态障碍物来切换路径规划机制,完成动态环境下的路径规划任务。仿真结果表明该算法有效可行,能在动态环境中为移动机器人实时规划出可行路径。

In order to solve the path planning problem of mobile robots in dynamic environment,twolayer path planning algorithm based on improved ant colony algorithm and MA-DQN algorithm is proposed.Static global path planning is accomplished by ant colony algorithm that improved the probabilistic transfer function and the pheromone updating principle;the traditional DQN algorithm structure is improved by using the memristor as the synaptic structure of neural network,and then completed the local dynamic obstacle avoidance of the mobile robot.The path planning mechanism is switched according to whether there are dynamic obstacles within the sensing range of the mobile robot,so as to completed the path planning task in the dynamic environment.The simulation results show that the algorithm can effectively plan a feasible path for mobile robots in a dynamic environment in real time.