混线生产系统典型干扰情境下基于仿真优化的批调度方法

项目名称： 混线生产系统典型干扰情境下基于仿真优化的批调度方法

项目编号： No.71502026

项目类型： 青年科学基金项目

立项/批准年度： 2016

项目学科： 管理科学

项目作者： 刘锋

作者单位： 东北财经大学

项目金额： 19万元

中文摘要： 流水-单元混线生产系统中市场需求、生产设备或装配工人相关的典型干扰事件中断了系统的正常运营，恶化调度性能的同时造成系统扰动。面对干扰事件有效应对的现实需求，本项目采用仿真优化从干扰管理的视角研究产品批次调度计划的制定。首先针对干扰事件的影响，提出系统扰动的度量方法。基于单元和流水工序结构、装配工人学习效应和批次安装时间等生产因素，构建考虑市场需求相关动态批次到达、生产设备相关随机故障、装配工人相关随机批次处理时间的批调度模型。通过模型性质分析，设计基于仿真优化的多目标进化算法，建立适应度值的代理模型从而降低运算代价，建立分布估计模型从而权衡广度搜索和深度搜索。最后通过数值实验和应用检验为混线系统生产管理提供政策启示。本项目将干扰管理研究延伸至流水-单元系统，不仅在理论上提出典型干扰情景下批调度的生产管理方法，也为实施混线生产方式的企业应对典型干扰事件提供理论、方法和技术支持。

中文关键词： 生产计划与调度；干扰管理；混线生产系统；仿真优化；进化计算

英文摘要： In hybrid line-seru production system, market-, production equipment-, and shop-floor worker-related disruptions interrupt the normal system operations, degrading schedule performance as well as causing system deviations. To address the practical need of effectively handling disruptions, in this project we use simulation-based optimization approach to study batch schedule planning from the perspective of disruption management. First, system deviation measuring methods are proposed for disruptions. Then based on production factors such as hybrid seru-line structure, worker’s learning effect and batch setup times, batch scheduling models are built up. And market-related dynamic batch arrival, equipment-related random breakdown, and worker-related random batch processing duration are considered respectively. By analyzing model properties, we design simulation-based multi-objective evolutionary algorithm, with surrogate model for fitness estimation to reduce computational burden and distribution estimation model to balance between exploration and exploitation. Finally policy implications for line-seru production management are proposed through numerical and application verifications. In this project, we extend researches on disruption management into hybrid line-seru system. From the theoretical aspect we establish production management methods for batch scheduling under typical disruption scenarios, whereas from the practical aspect we provide theory, method and technology support for enterprises adopting line-seru production system.

英文关键词： production planning and scheduling;disruption management;hybrid line-seru production system;simulation-based optimization;evolutionary computation