针对传统的粒子群算法易陷入局部最小，且算法后期的粒子速度下降过快而失去搜索能力等缺陷，本文提出了一种基于混沌思想的新型粒子群算法。该算法通过生成混沌序列的方式产生惯性权重取代传统惯性权重线性递减的方案，使粒子速度呈现多样性的特点，从而提高算法的全局搜索能力；根据算法中粒子群体的平均粒子速度调节惯性权重，防止粒子速度过早降低而造成的搜索能力下降的问题；最后通过引入粒子群算法系统模型稳定时惯性权重和加速系数之间的约束关系，增强了粒子群算法的局部搜索能力。对比仿真实验表明，本文所提改进的混沌粒子群算法较传统粒子群算法具有更好的搜索性能。

A modified particle swarm optimization algorithm is proposed which aims to solving the flaws of easy plunging into local optimum  and losing search ability in the last period for the fast particle velocity decrease. The paper introduces chaos mapping into the particle swarm optimization instead of the linear reduction inertia weight,and  prevents the velocity decrease early,the inertia weight is regulated according to the average particle velocity.In the last period of the algorithm,the constraint relation between the acceleration coefficient and the inertia weight is used to improve the local search ability. Simulations show that the search performance of the proposed method is much better than the traditional PSO algorithm.