一种自适应惯性权重粒子群算法及其仿真的研究

An Adaptive Inertia Weight Particle Swarm Algorithm And Simulation

对于传统粒子群算法容易陷入局部最优的问题,本文提出了一种自适应惯性权重粒子群算法。针对粒子群的早熟收敛程度设计了一个指标,使粒子权重根据早熟收敛度来自适应的调整,保证粒子群在全局寻优的过程中能快速收敛,避免陷入局部最优解。对本算法进行仿真,用于经典函数的优化,结果表明:本文算法能获得全局最优解,且快速收敛。

For the local optimum problem of the traditional PSO algorithm, an adaptive inertia weight particle swarm optimization is proposed in this paper. For premature convergence degree of particle swarm a index was designed, make sure that the particle weight can be adaptively adjusted according to the premature convergence and particle swarm have the ability of global optimum and fast convergence, avoid falling into local optimal solution. This algorithm was simulated and used for the optimization of the classical function. The results show that the algorithm can obtain the global optimal solution and fast convergence.