不确定可重入定点装配车间集成生产计划与调度  

Integrated optimization of production planning and scheduling for fixed-position assembly workshop with uncertain re-entrance

作　　者：蒋南云[1,2,3] 严洪森[1,2] 

机构地区：[1]东南大学自动化学院,江苏南京210096 [2]东南大学复杂工程系统测量与控制教育部重点实验室,江苏南京210096 [3]南京工业大学经济与管理学院,江苏南京210009

出　　处：《计算机集成制造系统》2017年第3期584-598,共15页Computer Integrated Manufacturing Systems

基　　金：国家自然科学基金重点资助项目(61673112;60934008);中央高校基本科研业务费资助项目(2242014K10031);江苏高校优势学科建设工程资助项目;江苏省教育厅高校哲学社会科学研究资助项目(2016SJB630025)~~

摘　　要：为满足工厂—车间一体化管理需求,研究了不确定可重入定点装配车间生产计划与调度集成优化问题。在分析车间装配特点的基础上,利用期望值描述不确定可重入情况,建立了双层生产计划与调度集成优化随机期望值模型,上层为能力约束的生产计划模型,下层为不确定可重入定点装配车间调度模型。提出了一种具有双层结构的交替迭代式改进遗传算法,上层用精英遗传算法求解生产计划,代入下层后采用基于随机模拟技术的遗传模拟退火算法求解生产调度,然后将调度结果返回上层重新求解新计划,如此不断交替迭代以实现计划与调度的同时优化。通过算例仿真验证了模型及算法的有效性。为制定不确定可重入定点装配车间生产计划与调度提供了一种合理可行的方法。To meet the demand of integration management between plants and workshops,the integrated optimization of production planning and scheduling for fixed-position assembly workshop with uncertain re-entrance was studied.Based on analyzing the assembly characteristic of workshop,the value of expectation was used to describe the uncertainty of re-entrance,and a bi-level stochastic expected value model of integrated production planning and scheduling was presented,in which the upper-level was a production planning model with capacity restriction and the lower-level was a scheduling model of fixed-position assembly workshop with uncertain re-entrance.An alternant iterative method with a bi-level structure by modified genetic algorithm was proposed.Elite genetic algorithm was applied to solve the problem of production planning in the upper-level,and the genetic simulated annealing algorithm based on stochastic simulation was applied to solve the problem of scheduling in the lower-level.The production plan was obtained in the upper-level firstly and then put into the lower-level to get the schedule,and the schedule was then putinto the upper-level to get the new plan.By such alternant iterative process,the optimization of production planning and scheduling was achieved simultaneously.The result of simulation examples indicated that the proposed model and algorithm were effective,which provided a feasible method to make the production planning and scheduling for fixed-position assembly workshop with uncertain re-entrance.

关 键 词：不确定可重入 定点装配 生产计划与调度集成优化 期望值模型 遗传模拟退火算法 

分 类 号：TP278[自动化与计算机技术—检测技术与自动化装置]