Abstract: A Dynamic Diffusion Particle Swarm Optimization algorithm（DDPSO） is proposed to improve the poor search quality of the standard PSO for optimizing high-dimensional function and its application to mobile robot path planning.The dynamic adjustment number is introduced to adjust the running track of particle dynamically as well as to prevent the drawback of reduction of efficiency during searching，and it can improve the performance of optimization algorithms.Simulations results show that DDPSO has more powerful optimizing ability and higher optimizing precision in high-dimensional function optimization and motion robot path planning.

Key words: particle swarm optimization, high-dimensional function optimization, dynamic adjustment number, path planning

摘要： 针对粒子群算法搜索精度不高特别是对高维函数优化性能不佳问题，提出了一种动态扩散粒子群算法，并将其应用于移动机器人路径规划中。该算法通过引进动态调节数，动态地选择粒子的运行轨迹，阻止种群在演化过程中搜索效率降低的缺陷，提高算法的寻优性能。实验结果表明，该算法在处理高维函数优化及移动机器人路径规划方面具有更强的寻优能力及更高的搜索精度。

关键词: 粒子群算法, 高维函数优化, 动态调节数, 路径规划