A competitive particle swarm algorithm based on multiple swarm and strategy
The author proposes a multi-population and multi-strategy competitive particle swarm optimization(PSO)algorithm to address the issues faced by the standard PSO,such as falling into local optima,poor conver-gence,and low solution accuracy.In the new algorithm,each generation of particles is sorted based on fitness,divid-ed into different sub-populations,and updated using nonlinear logistic chaotic mapping weights,contraction fac-tors,and a hybrid Gaussian-Cauchy perturbation mechanism.The use of different particle update strategies bal-ances the global search and local exploitation capabilities of the algorithm,thus accelerating the convergence speed.Finally,the multi-population and multi-strategy competitive PSO algorithm is compared with the standard PSO and other optimization algorithms on 11 test functions.The results demonstrate that the new algorithm significantly outperforms the standard PSO in terms of escaping local optima and solution accuracy,and it has a faster conver-gence speed.It also exhibits significantly higher optimization capability and algorithm stability compared to other benchmark algorithms.
万方数据
维普
