柔性作业车间调度多策略果蝇优化算法研究  

Multi-strategy fruit fly optimization algorithm for flexible job shop scheduling

作　　者：吴迎晨 肖彪 赵正彩[1] 彭仕鑫 苏宏华[1] 朱夏林 WU Yingchen;XIAO Biao;ZHAO Zhengcai;PENG Shixin;SU Honghua;ZHU Xialin(College of Mechanical&Electrical Engineering,NUAA,Nanjing 210016,China;Nanjing Chenguang Group Co.,Ltd.,Nanjing 210012,China)

机构地区：[1]南京航空航天大学机电学院,南京210016 [2]南京晨光集团有限责任公司,南京210012

出　　处：《现代制造工程》2023年第5期22-30,44,共10页Modern Manufacturing Engineering

基　　金：国防基础科研项目(JCKY2019204B015)。

摘　　要：针对柔性作业车间调度问题,提出了融合多种策略的果蝇优化算法,以最小化最大完工时间为目标,生成合适的调度方案。在该算法中,采用反向学习策略进行种群初始化,改善初始解的质量。其次引入变邻域搜索算法作为果蝇的嗅觉搜索策略,提升局部搜索能力;在此基础上,提出能够增加种群多样性的位置重构策略,根据嗅觉搜索后种群的平均适应度值将果蝇动态地划分为先进子种群和普通子种群,普通子种群以先进子种群为目标进行位置重构,先进子种群相互为目标进行位置重构,避免搜索陷入局部最优。最后,设计正交实验探究变邻域搜索参数对算法性能的影响,并确定相关参数合适值;通过计算多个标准测试集基准算例,得到每个测试集的最优解及平均解,并与现有算法的计算结果进行对比,验证了改进策略的有效性及所提算法的优越性。Aiming at the flexible job shop scheduling problem,a fruit fly optimization algorithm integrating multiple strategies was proposed to generate an appropriate scheduling scheme with minimizing makespan.In this algorithm,the opposite learning strategy was used to initialize the population to improve the quality of the initial solution sets;and then a variable neighborhood search algorithm was adopted as the smell-based search strategy of the individuals to improve the local search ability.On this basis,a position reconstruction strategy that can increase the diversity of the population was proposed.According to the average fitness values of the individuals after the smell-based search,the fruit flies were dynamically divided into advanced subpopulation and ordinary subpopulation.Individual positions of the ordinary subpopulation were reconstructed with the advanced individuals as the target,and individual positions of the advanced subpopulation were reconstructed with each other as the target,so as to avoid the search falling into local optimization.Finally,orthogonal experiments were designed to explore the influence of variable neighborhood search parameters on the performance of the algorithm,and the appropriate values of each parameter were determined.By calculating several groups of benchmark testing sets,the optimal solutions and average solutions of each group of testing sets were obtained and compared with the calculation results of existing algorithms,which verifies the effectiveness of the improved strategies and the superiority of the proposed algorithm.

关 键 词：柔性作业车间调度 果蝇优化算法 变邻域搜索策略 重构策略 

分 类 号：TP186[自动化与计算机技术—控制理论与控制工程]