基于改进IGA的多品种变批量智能车间调度  

Intelligent workshop scheduling with multiple varieties and variable batches based on improved IGA

作　　者：刘晋飞[1] 刘乙涵 陈明[2] 黄华 LIU Jinfei;LIU Yihan;CHEN Ming;HUANG Hua(Sino-German College of Applied Sciences,Tongji University,Shanghai 201804,China;School of Mechanical Engineering,Tongji University,Shanghai 201804,China)

机构地区：[1]同济大学中德工程学院,上海201804 [2]同济大学机械能源工程学院,上海201804

出　　处：《现代制造工程》2025年第4期1-10,共10页Modern Manufacturing Engineering

基　　金：国家自然科学基金项目(71601144)。

摘　　要：针对多品种、变批量的高复杂度智能制造场景,频繁更换刀具、夹具及工装等情况造成的实际生产调度和理论生产调度脱节的问题,定义了两个参量,即机器准备时间(Machine Preparation Duration,MPD)和机器加工系数(Machine Processing Coefficient,MPC),以最小化最大完工时间、机器总时间负荷和机器总准备时间为目标函数,建立了引入MPC参数的多品种、变批量智能车间调度数学模型;设计了融合非支配排序遗传算法(Non-dominated Sorting Genetic Algorithm-Ⅱ,NSGA-Ⅱ)和免疫遗传算法(Immune Genetic Algorithm,IGA)的非支配免疫遗传算法(Non-dominated Sorting Immune Genetic Algorithm-Ⅱ,NSIGA-Ⅱ)来求解此类问题。该算法采用多种方式进行初始化,提出了一种综合考虑非支配排序和目标函数值大小的得分策略来筛选优秀个体,同时为了提高种群的多样性,引入种群分层和自适应交叉突变的策略。最后,通过多组对比实验验证了该算法的有效性以及在探索最优解时具有稳定性好、解质量高等优点。In view of the problem of disconnection between actual production scheduling and theoretical production scheduling due to frequent replacement of tools,fixtures,etc.in high-complexity intelligent manufacturing scenarios with multiple varieties and variable batches,it defines two coefficients,Machine Preparation Duration(MPD)and Machine Processing Coefficient(MPC),constructs an intelligent workshop scheduling mathematical model with multiple varieties and variable batches that introduces MPC parameters and takes minimizing maximum processing machine duration,the total machine load time and the total machine preparation time as the objective functions,designs a new solving algorithm Non-dominated Sorting Immune Genetic Algorithm-Ⅱ(NSIGA-Ⅱ)that combines Non-dominated Sorting Genetic Algorithm-Ⅱ(NGSA-Ⅱ)and Immune Genetic Algorithm(IGA).This algorithm uses a variety of methods for initialization,and proposes a scoring strategy that comprehensively considers non-dominated sorting and the size of the objective function value to select good individuals.At the same time,in order to improve the diversity of the population,it introduces the strategy of population stratification and adaptive cross-mutation.Finally,multiple sets of comparative experiments verify the effectiveness of this algorithm and its advantages of good stability and high solution quality when exploring optimal solutions.

关 键 词：机器准备时间 非支配排序算法 免疫遗传算法 智能车间调度 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] TH164[自动化与计算机技术—控制科学与工程]