基于时空域模型分解策略的流程企业级协同优化方法研究

项目名称： 基于时空域模型分解策略的流程企业级协同优化方法研究

项目编号： No.61304211

项目类型： 青年科学基金项目

立项/批准年度： 2014

项目学科： 自动化技术、计算机技术

项目作者： 郑松

作者单位： 杭州电子科技大学

项目金额： 24万元

中文摘要： 实施企业级优化（EWO，Enterprise-wide Optimization）是提高企业核心竞争力的重要举措，也是近年来系统优化领域新兴的研究课题。目前EWO问题的模型规模复杂，不利于求解；当前处理EWO问题的分散决策方法，会造成各DMU（Decision Making Unit）之间平等性差；另外，优化多模型复杂系统的CO（Collaborative Optimization）算法鲁棒性弱。 本项目研究从时空域角度对EWO模型分解的方法，减少DMU数量，降低单个DMU模型复杂度；基于CO方法论，应用动态松弛法松弛各DMU间的一致性约束式(物质流、信息流)，将EWO优化问题分解为系统级、学科级（DMU级）两级优化问题；最后，分析CO的几何特性，研究不同区域内松弛量与初始点所在区域和学科一致性的动态关系，提出一致性约束松弛量的自适应调整方法。

中文关键词： 协同优化；企业级优化；决策单元；模型分解；

英文摘要： The implementation of enterprise-wide optimization (EWO) is an important measure to improve the core competitiveness of enterprises. It is also the emerging research topics of system optimization in recent years. The EWO model is complex and difficult to obtain optimal solution. There is inequality between the decision making unit (DMU) in the decentralized decision-making method for EWO. In addition, The robustness of collaborative optimization (CO) is weak, which is fit for the optimization of multi-model complex system. The method decomposition of EWO model, collaborative optimization method for the EWO and the adjustment method in adaptive constraint relaxation quantity will be studied in this project. The temporal and spatial model decomposition is used to Reduce the number and the complexity of DMU. CO method is used to solve the EWO problem, apply the dynamic relaxation method to relax the consistency constraints (material flow, information flow) among each DMU, then the EWO optimization problem is decomposed into the two level optimization problem composed by system-level and discipline level (DMU level). Finally, The project will study the dynamic relationship among the slack factor, initial point and disciplinary consistency, and the adaptive adjustment of the constraint relaxation method will be prese

英文关键词： Collaborative Optimization；Enterprise-wide Optimization；Decision Making Unit；Model Decomposition；