Abstract: This paper proposes a novel Modified Adaptive Particle Swarm Optimization（MCAPSO） algorithm based on cloud theory to improve the optimum speed and performance of the PSO algorithm. The particles are divided into three groups based on the fitness of the particle in order to adopt different inertia weight generating strategy and evolutionary strategy and effective balance between the local and global search ability is achieved. This paper chooses five reference functions to have a test and compares the results with other PSO algorithms. The simulation results verify the effectiveness of this approach.

Key words: Particle Swarm Optimization（PSO）, adaptive inertia weight based on cloud theory, evolutionary strategy

摘要： 为了提高粒子群算法的寻优速度和精度，提出一种改进的云自适应粒子群算法（MCAPSO）。算法中根据粒子适应度值把种群分为三个子群，分别采用不同的惯性权重生成策略和进化策略，普通子群粒子采用云自适应惯性权重，有效地调整了算法的全局与局部搜索能力。选取了五个基准函数进行测试，与其他PSO算法作了比较。仿真结果表明该方法是有效的。

关键词: 粒子群算法, 云自适应惯性权重, 进化策略