微粒群算法参数的理论分析

微粒群算法是近年来兴起的一种智能优化算法，而算法参数是影响算法性能和效率的关键。用基于常系数非齐次差分方程求解的分析、基于动态系统理论的分析和基于离散系统稳定判据的分析三种不同的方式对微粒的位置和速度两个变量进行了深入理论分析，最终得出了一个共同的结论，即保证微粒收敛的参数取值区域约束在一个直角梯形的内部，这将对算法的实际应用起到很重要的作用。

The Parameters Theoretical Analysis of Particle Swarm Optimization Algorithm

In recent years, particle swarm optimization （PSO） was a kind of intelligent optimization algorithms, and the algorithm parameters was a key affect to algorithm performance and efficiency. The location and speed of particles were deeply analyzed by three different ways of non-homogeneous differential equation solving based on constant coefficient, dynamic systems theory and stability criterion for discrete systems. Finally work out a common conclusion that is the parameter values ensure convergence of particles bound in the internal region of a fight-angle trapezoidal.