应用混沌烟花算法求解置换流水车间问题

APPLYING CHAOTIC FIREWORKS ALGORITHM IN SOLVING PERMUTATION FLOW SHOP PROBLEM

改进烟花算法求解置换流水车间问题。用最大位置法编码,将连续变量映射到离散空间。引入动态半径因子,平衡局部搜索与全局搜索。精英个体混沌搜索,进一步挖掘个体信息。用锦标赛策略替代原有的选择算子,群体中的优良个体被选择的概率增大。通过正交实验选择合适参数,求解Car类和Rec类基准问题。与基本烟花算法、萤火虫算法和粒子群算法的对比实验说明,改进后的混沌烟花算法在寻优率、寻优速度等上具有一定的优势,是求解置换流水车间问题的有效工具。

We improved the fireworks algorithm to solve PFSP. By encoding with maximum position method,we mapped the continuous variables onto discrete space. To strike a balance between global searching and local searching,we introduced dynamic radius factor. We further mined the individual information with elite individual chaotic search. We replaced original selection operator with champion contest strategy,and as a result,the excellent ones in population could be selected at a higher rate of probability. We chose right parameters through orthogonal experiment for solving the benchmark problems of Car class and Rec class. Comparative experiments on basic fireworks algorithm,firefly algorithm and particle swarm optimisation illustrated that the improved chaotic fireworks algorithm has certain advantage over other algorithms in searching rate and searching speed and is an effective tool of solving permutation flow shop problem.