Abstract: Concerning the fact that the standard Particle Swarm Optimization(PSO) algorithm has the problem of population diversity lose and premature convergence, using the feature of independent individual behavior of social animals in nature for reference, this paper proposes the concept of individual state. A Particle Swarm Optimization(PSO) algorithm based on individual state and state transition is proposed and tested with several typical benchmark functions. The result indicates that the algorithm is significantly superior to standard PSO in performance of optimization. Compared with other improved algorithms, it is also excellent in performance of optimization.

Key words: Particle Swarm Optimization(PSO) algorithm, individual state, global optimization

摘要： 针对标准粒子群算法的种群多样性丧失和算法早熟收敛问题，借鉴自然界中群居动物个体行为的独立性特征，提出粒子的个体状态概念，给出一种基于微粒个体状态和状态迁移的粒子群优化算法。对典型函数测试结果的比较表明，改进后算法的寻优能力明显高于标准粒子群算法。与其他改进算法相比，该算法的寻优能力也较强。

关键词: 粒子群优化算法, 个体状态, 全局优化

CLC Number: 

• TP301.6