一种非线性动态自适应惯性权重PSO算法  

A Nonlinear Dynamic Adaptive Inertial Weight Particle Swarm Optimization

作　　者：王生亮 刘根友[1] WANG Sheng-liang;LIU Gen-you(Institute of Geodesy and Geophysics,Chinese Academy of Sciences,Wuhan Hubei 430077,China;College of Earth and Planetary Sciences,University of Chinese Academy of Sciences,Beijing 100049,China)

机构地区：[1]中国科学院测量与地球物理研究所,湖北武汉430077 [2]中国科学院大学地球与行星科学学院,北京100049

出　　处：《计算机仿真》2021年第4期249-253,451,共6页Computer Simulation

基　　金：国家重点研发计划项目(2016YFB0501900);国家自然科学基金资助项目(41621091,41774017)。

摘　　要：传统的时变惯性权重粒子群优化算法对于求解一般的全局最优问题具有良好的效果,而对于复杂高维的优化问题易陷入局部收敛、存在早熟等缺点。针对以上存在的缺点,提出了种群进化离散度的概念,并考虑Sigmoid函数在线性与非线性之间较好的平衡性能,给出一种非线性动态自适应惯性权重的粒子群优化算法。该算法充分考虑进化过程中种群粒子之间进化差异,自适应地赋予不同的惯性权重因子,满足粒子群优化算法在不同进化时期对全局探索和局部开发能力的需求,仿真实例测试结果验证了该算法的有效性。The traditional time-varying inertia weight particle swarm optimization algorithm has a good effect on solving general global optimization problems, but for complex high-dimensional optimization problems, it is easy to fall into local convergence and has premature shortcomings. This paper proposed the concept of population evolution dispersion. In addition, considering the better balance between linearity and nonlinearity of Sigmoid function, a particle swarm optimization algorithm with nonlinear dynamic adaptive inertial weights was proposed. The algorithm fully considers the evolutionary differences between population particles in the evolutionary process, and different inertia weight factors can be adaptively assigned to meet the requirements of the particle swarm optimization algorithm for global exploration and local development ability in different evolutionary periods. The simulation results show that the proposed algorithm is effective.

关 键 词：粒子群优化 动态自适应 惯性权重 进化离散度 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]