复杂分布式制造系统的调度优化理论及方法研究

In order to get lower production cost, decrease manufacturing period and achieve higher product quality, many enterprises have changed their single factory production system to distributed production system which has multi-factories with similar functions. This project will study three complex disributed scheduling problems including distributed flexible flowshop scheduling problem, distrbuted hybrid fabrication-assembly flowshop scheduling problem and distributed fabrication-assembly-diferentiation flowshop scheduling scheduling problem. At first, the mixed integer programming (MIP) models will be estabilished for these problems. Then, the properties of the optimal soultions in the special- and general cases will be presented. For small size problems, this project will develop the branch and bound algorithms based on the properties of the problems, whereas for medium- and large size problems, this project will construct heuristics and hybrid swarm intelligent algorithms based on problem properties and statistics information. The empasis will be located on the encoding,decoding,integration of problem properties and metaheuristics,hybrid metaheuristics, and local search methods. Aimed at evoluating the performances of proposed algorithms, several lower bounds algorithms will be developed. At last, with a large amount of experiments,we verify the effectiveness and efficientness of the proposed MIP models and algorithms. This research will not only enrich the theory and methods of the distributed scheduling problem, but also instruct the production operations management for the distributed manufacturing enterprises.

为降低生产成本、减少生产周期并提高产品质量，越来越多的制造企业从单工厂制造模式转变为多工厂分布式制造模式。以复杂分布式制造模式为背景，本项目主要研究三类复杂分布式制造系统调度问题，包括：分布式柔性流水调度问题，分布式柔性流水加工-装配调度问题以及分布式加工-装配-分化调度问题。首先建立三类复杂分布式调度问题的混合整数规划模型。随后研究一般情形下和特殊情形下问题最优解的性质。算法方面，主要研究解决相应小规模调度优化问题的分支定界算法，解决中大规模调度优化问题的快速启发式算法以及基于问题性质和统计信息的混合元启发式算法。在此算法框架基础上，研究各种混合智能优化算法的编码解码方式，问题性质与元启发式算法的集成方式，不同智能优化算法之间的混合方式，通过大量算例研究模型及算法的有效性。该项研究不仅可以丰富分布式制造系统调度优化理论及方法，而且对分布式制造企业的生产运作管理有着重要的指导意义。

为快速适应动态多变的市场需求，很多制造企业都从传统的单工厂模式向分布式多工厂模式转变。而分布式制造系统的调度优化则可以显著提升资源的分配和使用效率，为企业增加利润。该项目重点研究了分布式制造环境下的柔性加工调度问题，柔性加工-装配调度问题以及加工-装配-分化调度问题。首先建立了三类复杂分布式调度问题的数学模型，包括数学规划模型和Petri网模型。随后给出了一般情形下和特殊情形下问题最优解的性质和无死锁控制策略。算法方面，提出了解决相应小规模调度优化问题的分支定界算法，解决中大规模调度优化问题的快速启发式算法以及基于问题性质和统计信息的混合元启发式算法。提出了多种混合智能优化算法的编码解码方法，无死锁编码解码方法，问题性质、无死锁策略与智能优化算法的集成方法，通过大量算例验证了模型及算法的有效性。该研究不仅丰富了分布式制造系统调度优化理论及方法，而且对分布式制造企业的生产运作管理有着重要的指导意义，有望在电子、汽车等分布式离散制造行业得到推广应用。