基于改进离散粒子群优化算法的路径规划

Path Planning Based on Improved Discrete Particle Swarm Optimization Algorithm

针对传统粒子群算法进行机器人路径规划时容易陷入局部最优的问题,提出一种基于反向学习机制的改进离散粒子群优化算法。算法分两阶段对粒子种群引入反向学习机制:第一次在种群初始化时,对全部粒子引入反向学习机制,选择适应度值较低的粒子作为初始种群,提高初始解的质量;第二次在迭代过程中,以概率α对粒子种群引入反向学习机制,增加粒子种群的多样性。使用改进惯性权重,提出一种新型自适应余弦变化公式更新学习因子,使算法在运行过程中达到更佳的搜索性能。利用Matlab R2014a平台进行仿真,结果表明,改进算法可以在无碰撞条件下规划出一条从初始点运动到目标点的最短路径,且收敛速度和精度均得到提高。

An improved discrete particle swarm optimization algorithm based on reverse learning mechanism is proposed to solve the problem that the traditional particle swarm optimization algorithm is prone to fall into local optimization in robot path planning. The algorithm introduces a reverse learning mechanism to the particle population in two stages. In the first stage, a reverse learning mechanism is introduced for all particles during population initialization, and particles with a lower fitness value are selected as the initial population, which can improve the quality of the initial solution. The second stage is the iteration process, a reverse learning mechanism is introduced to the particle population with probability αto increase the diversity of the particle population. Using improved inertia weights, a new adaptive cosine variation formula is proposed to update the learning factors so that the algorithm can achieve better search performance during operation. The Matlab R2014 a platform was used for simulation experiment. The results demonstrate that the improved algorithm can plan the shortest path from the initial point to the target point under collision-free conditions, and the convergence speed and precision are both improved.