不确定条件下生产调度问题的三阶段决策方法  

Three-stage decision method for production scheduling under uncertainty

作　　者：丁然[1] 李歧强[1] 郭庆强[1] 刘圆圆[1] 

机构地区：[1]山东大学控制科学与工程学院,山东济南250061

出　　处：《控制理论与应用》2008年第6期1158-1162,1166,共6页Control Theory & Applications

基　　金：山东省优秀中青年科学家奖励基金资助项目(2007BS05014);国家863计划项目(2007AA04Z157).

摘　　要：提出了一种可综合考虑多种不确定因素的生产调度问题三阶段决策方法.首先分析了不确定条件下的生产调度的决策过程,根据过程的特点将其分为基本调度、在线调整和补偿三个阶段;对变量进行分类,从而与3个阶段相对应;分别建立了基于情景分析以及子问题最(近)优解的两种三阶段调度数学模型,并给出了变量的处理方法和阶段性的模型求解方法,来作为不同阶段的决策依据;最后以化工批处理过程的短期调度为例,说明了三阶段调度建模和决策过程的实现方法.A novel three-stage decision method considering multiple sorts of uncertainties for production scheduling is proposed. Firstly, the decision process for production scheduling under uncertainties is analyzed and divided into three stages, which are basic scheduling, on-line adjusting and recourse. Secondly, the variables are divided into three sorts according to the three stages. Then, two kinds of three-stage mathematical models based on scene-analysis and （approximate） optimal solution are presented, along with the methods for handling variables and solving the problem. These methods for solving problem by stages constitute the basis of the decision making. To illustrate the process of modeling and decision, a chemical batch process is considered as an example.

关 键 词：不确定 生产调度 决策 多阶段 批处理 

分 类 号：TB114.1[理学—运筹学与控制论]