基于动态自适应变参的粒子群优化算法

Particle Swarm Optimization Algorithm Based on Dynamic Adaptive Variable Parameters

在剖析参数对粒子群算法效率和寻优能力影响的基础上,提出了动态自适应时变参数优化的粒子群算法,根据迭代过程对算法中的惯性权重、加速因子进行对应的非线性自适应调整,同时引入动态自适应的控制因子避免寻优过程中粒子因速度远离全局最优位置,并将运用变参优化的粒子群算法应用于三维和多维函数寻优。实验结果表明:动态自适应变参优化的粒子群算法比标准粒子群算法有更好的算法效率和寻优能力,同时在解决多维优化问题上亦具有出色表现。运用结果证明基于动态自适应时变参数优化的粒子群算法较普通的粒子群算法更有优越性。

Based on the analysis of the influence of parameters on the efficiency and optimization ability of particle swarm optimization algorithm,a dynamic adaptive variable parameter strategy optimizated particle swarm optimization algorithm is proposed.The inertia weight and acceleration factor in the algorithm are non-linearly adaptively adjusted according to the iteration process.At the same time,the dynamic adaptive control factor is introduced to avoid the particles from moving away from the global optimal position due to the speed in the optimization process,and the particle swarm algorithm using variable parameter optimization is applied to the optimization of three-dimensional and multi-dimensional functions.The experimental results show that the dynamic adaptive variable parameter optimization particle swarm optimization algorithm has better algorithm efficiency and optimization ability than the standard particle swarm optimization algorithm,and it also has excellent performance in solving multi-dimensional optimization problems.The results show that the PSO algorithm based on dynamic adaptive time-varying parameter optimization is superior to the ordinary PSO algorithm.