Abstract: The penalty function is used to change the constrained problem into the unconstrained problem in path planning of robots. By using the randomicity and stable tendentiousness characteristics of cloud model, an adaptive strategy for varying parameters of Particle Swarm Optimization（PSO） theory is introduced based on cloud model. So an improved Particle Swarm Optimization（PSO） algorithm is constructed and applied to path planning of robots. By adopting different inertia weight generating methods in different groups, the searching ability of the algorithm in local and overall situation is balanced effectively. And it does not only improve the convergence speed, but also maintains the diversity of the population. The feasibility and effectiveness are proved by the comparative results of the simulation experiments. And also the algorithm can be achieved simply and has fast convergence rate.

Key words: cloud model, Particle Swarm Optimization（PSO） algorithm, path planning, adaptive varying parameters

摘要： 利用罚函数将机器人路径规划有约束优化问题转换为无约束优化问题。利用云模型既有随机性又有稳定倾向性的特性，引入基于云模型理论的自适应参数策略，构造出一种改进的粒子群（PSO）算法，并应用于机器人路径规划问题。在不同的子群采用不同的惯性权重生成方法，有效地平衡了算法的局部和全局搜索能力，提高了种群的多样性和算法的收敛速度。仿真结果对比验证了该算法的可行性和有效性，且实现简单、收敛速度快。

关键词: 云模型, 粒子群算法, 路径规划, 自适应参数调整