考虑节能的改进多目标樽海鞘群算法TFT-LCD面板阵列制程调度问题  

IMSSA for TFT-LCD Panel Array Process Scheduling Problem Considering Energy Saving

作　　者：姚远远[1] 叶春明[1] YAO Yuanyuan;YE Chunming(Business School,University of Shanghai for Science and Technology,Shanghai,200093)

机构地区：[1]上海理工大学管理学院

出　　处：《中国机械工程》2019年第24期2994-3003,共10页China Mechanical Engineering

基　　金：国家自然科学基金资助项目(71840003);上海理工大学科技发展基金资助项目(2018KJFZ043)

摘　　要：TFT-LCD面板生产的阵列制程是可重入混合流水车间调度问题,采用一种改进多目标樽海鞘群算法对其进行优化求解。构建以最大完工时间、总拖期时间和总耗能为优化目标的数学规划模型;针对该问题结构特点,对基本多目标樽海鞘群算法进行了一系列改进操作,包括基于升序排列的随机键编码、PS方法解码、基于Lévy飞行的领导者个体位置更新方式,以及外部档案中非支配个体的变邻域搜索操作,并采用田口方法进行算法参数设置;最后通过对基准算例的数值实验,将改进多目标樽海鞘群算法与基本多目标樽海鞘群算法、多目标粒子群优化算法、快速非支配排序遗传算法进行对比,实验结果表明了改进多目标樽海鞘群算法的有效性。Array process of TFT-LCD panel manufacturing was a reentrant hybrid flow shop scheduling problem,and an effective IMSSA was proposed to solve the problem.Firstly,a multi-objective mathematical programming model with makespan,total tardiness,and total energy consumption criteria was formulated.Secondly,considering the characteristics of the problem,a series of improvements were made based on basic multi-objective salp swarm algorithm(MSSA),which included ranked order value based random key encoding,PS decoding methods,updating the position of the leading salps based on Lévy flight,and embedding a variable neighborhood search strategy in external archive.Influences of parameter setting were investigated by means of Taguchi method.Finally,the proposed IMSSA was compared with the basic MSSA,multi-objective particle swarm optimization(MOPSO)and fast nondominated sorting genetic algorithm(NSGA-Ⅱ)based on several benchmarking instances.Experimental results show the effectiveness of IMSSA.

关 键 词：可重入混合流水车间调度 改进多目标樽海鞘群算法 阵列制程 节能 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]