基于多智能体局部寻优的粒子群优化算法

为了提高粒子群算法的收敛速度和全局寻优能力，用多智能体遗传算法对粒子群算法当前搜索到的全局极值进行局部寻优．用搜索到的更好的解在下一次迭代中引导粒子进行搜索从而获得更快的收敛速度和更好的全局收敛性。对函数优化和神经网络训练的仿真实验表明．此算法能更快的收敛到全局最优解。

Particle Swarm Optimization Algorithm Based on Partial Search of Multi-Agent

To raise convergent speed and global search capacity of particle swarm optimization(PSO),this paper optimized the global extremum which currently searched by PSO on the partial with the Multi-Agent genetic algorithm(MAGA),the better solution searched made the PSO obtain the quicker convergence rate and the better global astringency in the next iteration.The simulation results of the function optimization and neural network training show that this algorithm could find the optimal solution more rapidly.