基于多群体并行遗传算法的混流混合车间模糊调度研究  

Study of mixed-model hybrid shop fuzzy scheduling problem based on multi-populations parallel genetic algorithm

作　　者：胡恒[1] 鲁建厦[1] 李英德[1] 

机构地区：[1]浙江工业大学机械工程学院,浙江杭州310032

出　　处：《浙江工业大学学报》2012年第5期554-558,共5页Journal of Zhejiang University of Technology

基　　金：国家自然科学基金资助项目(7097118);浙江省重大专项和优先主题计划基金资助项目(2009C11164);浙江省自然科学基金资助项目(LY12E05021)

摘　　要：为了有效的处理混流混合车间的物流堵塞问题,同时解决加工时间和交货期等参数的不确定性问题,提出了基于缓存区存量最低的混流混合车间的模糊优化数学模型.通过求解这个模型,能够实现在生产需求约束条件下,零件的缓存量最低,从而达到物流顺畅.此外,在模型的求解方法上,给出了一种收敛过程是阶段性、种群个体迁入、迁出是筛选性的多群体并行遗传算法,该算法能有效的保留种群的最优个体并根据收敛代数调整种群规模,从而增强解的全局搜索能力.实验结果表明:模型是合理的,算法是有效的,且求解结果不仅能满足客户的需求,而且能够有效地降低中间库存量,从而保证生产稳定、连续的进行.To solve the problem that high buffer of the mixed-model shop and mixed-flow production leading to logistics jam, meanwhile tackle the uncertain influences, a mixed-model hybrid shop fuzzy scheduling model that based on minimum buffer has been proposed in this paper. After solving this model problem, it can realize that buffer can be minimum and shop＇s logistics can be fluent. In addition, a new given multiple population parallel genetic algorithm that whose process has multiple stage and can quickly search the optical solution is proposed. The simulation states that the algorithm and the model is effective.

关 键 词：模糊 遗传算法 混流混合 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]