基于神经过程-粒子群算法的移动机器人路径规划

Path Planning of Mobile Robot Based on Neural Process-particle Swarm Optimization

针对传统粒子群算法在移动机器人路径规划过程中早熟引起的局部最优问题,将运动过程预测思想集成到粒子群优化算法中,构造神经过程-粒子群混合算法。主要思路是在粒子群个体进行下一次迭代时,利用神经过程预测个体位置,增加了迭代后期粒子群体的多样性,避免过早陷入局部最优,从而提高算法优化能力。实验结果显示,改进算法用于解决机器人路径规划问题,整体性能优于传统的粒子群优化算法。

Aiming at the local optimal problem caused by precocious particles in the path planning process of mobile robot by traditional Particle Swarm Optimization(PSO)algorithm,a hybrid neural process-PSO algorithm was constructed by integrating the idea of motion process prediction into PSO.The main idea is that in the next iteration of particle swarm individuals,the neural process is used to predict the individual location,increase the diversity of particle swarm at the later stage of the iteration,and avoid falling into local optimization too early,so as to improve the optimization ability of the algorithm.The improved algorithm is used to solve the robot path planning problem.The experimental results show that the proposed neural process-particle swarm optimization(PSO)has better path planning ability and better comprehensive performance than the traditional PSO.