集成生产与配送的供应链调度及其混合智能决策模型研究

项目名称： 集成生产与配送的供应链调度及其混合智能决策模型研究

项目编号： No.71302134

项目类型： 青年科学基金项目

立项/批准年度： 2014

项目学科： 管理科学

项目作者： 郭钊侠

作者单位： 四川大学

项目金额： 19万元

中文摘要： 按单制造（MTO）型制造企业在我国国民经济中占有非常重要的地位。本项目以全球化背景下兼具多工厂、多运送方式、多种订单大小等运作特征的典型MTO型制造企业为研究对象，针对其所面临的集成生产与配送的供应链调度问题的实际需求，从总公司（多工厂）、分工厂（单工厂）及不确定性环境三个层面分别对该类问题进行研究。基于计算智能与仿真技术，发展有效的混合智能决策模型，为该类问题提供有效的方法论。首先，将自适应和声搜索与文化基因算法集成，建立高效的混合智能寻优过程；在此基础上，再分别基于帕累托优化概念和蒙特卡洛仿真技术处理多个供应链调度目标和生产与配送不确定性；利用仿真的方法评价候选解并对所提出的决策模型进行验证。本项目成果可丰富和扩展调度与优化、供应链管理、计算智能等领域的知识和理论，所提出的智能方法和模型可在复杂的调度和优化问题中得到广泛应用，对相关企业供应链性能的提升和优化决策理论的发展具有重要意义。

中文关键词： 集成调度；供应链调度；混合智能；文化基因优化；多目标优化

英文摘要： Make-to-order (MTO) manufacturing companies play a crucial role in China's national economy. This research addresses the supply chain scheduling (SCS) problem integrating production and distribution operations, faced by China's MTO manufacturing companies with realistic operations features such as multiple plants, multiple transportation methods and multiple order sizes. In accordance with the actual requirements of these companies in global manufacturing, this research investigates the SCS problems integrating production and distribution operations in 3 different manufacturing scenarios respectively, including controlling company (multi-plant), branch plant (single-plant), and uncertain environments. Effective hybrid intelligent decision-making models are developed based on computational intelligence and simulation techniques, for providing effective methodologies for the investigated SCS problems. A highly-efficient optimum-seeking process is established firstly by integrating memetic algorithm wtih self-adaptive harmony search; Pareto optimality concept and Monte Carlo simulation technique are then adopted to handle multiple SCS objectives and various uncertainties respectively based on this process; simulation methods are used to evaluate candidate solutions and validate the decision-making models developed.

英文关键词： Integrated Scheduling；Supply Chain Scheduling；Hybrid Intelligence；Memetic Optimization；Multi-objective Optimimzation