改进协同粒子群优化算法及其在FlowShop调度中的应用

An Improved Cooperative Particle Swarm Optimization and Its Application to Flow Shop Scheduling Problem

针对协同粒子群优化算法存在的停滞现象,提出了一种改进的协同粒子群优化算法。采用优化法的子群协作方式,既保证了收敛速率,又可以防止陷入局部最优。同时引入综合学习策略,增加种群的多样性,防止种群出现停滞现象。在此基础上,又加入了扰动机制,进一步避免算法陷入局部最优。采用该算法对3个经典函数进行测试,并将其应用于Flow Shop调度问题,仿真实验结果表明:新算法有效克服了停滞现象,增强了全局搜索能力,比基本协同粒子群优化算法的优化性能更好。

Aiming at the stagnation problem of the cooperative particle swarm optimization, this paper presents an improved cooperative particle swarm optimization. This proposed method adopts the cooperation principle of optimization algorithm, so it not only ensures the convergence rate, but also avoids plunging into local optimum. Moreover, both comprehensive learning and disturbing mechanism are introduced to strengthen the diversity of population and avoid the stagnation and plunging into local optimum. The new algorithm is tested by three typical functions and the flow shop scheduling problems, respectively. The simulation results show that the proposed algorithm can avoid the stagnation, improve the global convergence ability, and attain better optimization performance.