【作者】 熊峰；

【导师】 张忠桢； Ram Akella；

【作者基本信息】 武汉理工大学 ， 管理科学与工程， 2009， 博士

【摘要】 面对全球化的浪潮和越来越激烈的市场竞争,供应链之间的竞争压力越来越大。如何快速响应不确定性的市场需求,在降低制造和配送成本、缩短产品生产和配送时间的情况下,实现高效的供应链协同制造,成为当前供应链管理者面临的非常具有挑战性的重要课题。通过研究,管理者和研究人员认识到,必须对产品在供应链内的生产制造以及分销领域内的全过程进行协同优化,寻找从供应链整体的角度对供应链内资源进行优化控制的方法,对于提高供应链生产效率,从而提升供应链竞争力,充分发挥供应链制造生产模式的优越性,具有非常重要的意义。因此对供应链协同调度进行研究有着广泛的现实背景和重要的理论意义与学术价值。本文在对国内外供应链优化和供应链协同调度领域里的研究方法和手段进行了总结和归纳的基础上,运用定量方法研究了供应链协同调度模型和求解方法,阐述了供应链协同调度的建模及优化的思路,并针对几个典型问题进行了建模并给出了具体的算例以及算法。论文的主要创新工作包括:(1)整合了供应链内多阶段调度模式。本文考虑了生产与分销多个环节集成模式下的整体最优安排,将供应链生产的综合成本和时间作为优化的目标,并建立了相应的多目标决策优化调度模型(SCISM模型),在此基础上,考虑缺货成本下的供应链成本增加的情况,建立了一个缺货成本下的二次规划优化调度模型。最后考虑了集成调度下的一个两阶段的生产和配送环节下的集成调度模型(OADSIM模型)。(2)针对线性和非线性的调度优化模型,运用旋转算法进行求解,旋转算法比较适合相对较大规模问题的计算,通过求解多目标线性规划和二次规划调度模型的算例,计算效果非常理想,均在较短时间内找到最优解。(3)把分组嵌套的遗传算法运用到大型供应链集成调度问题的求解中来。在实际中,由于供应链集成环境下,考虑生产厂家和配送过程的集成优化,既是一个指派问题,又具有排序问题的特征,这给模型的求解带来非常大的计算难度,这里给出了一种嵌套的遗传算法,在此基础上找寻集成调度问题的最优调度方案。(4)提出了基于多智能体的WEB集成调度系统框架。考虑供应链集成环境下基于多智能体的制造子系统和供应网络子系统的功能框架,结合实际需求,提出了基于Token的竞标型协商机制。该机制能够实现成员内基于整体资源调度的协商的快速实现。更多还原

【Abstract】 It has become a vital and challenging issue for the current supply chain management to study how to respond to uncertain market demand quickly, and fulfill an efficient supply chain collaborative manufacturing while reducing the manufacturing and distribution costs, reducing production and distribution of time. Through the research, Managers and researchers realize the need for the entire process of products manufacture both in the supply chain and the field of distribution, and to have a collaborative optimization. To find out a technique which optimize the resources of supply chain from the overall supply chain. Afterwards enhance the supply chain Production efficiency and the competitiveness of the supply chain. Give full play to the superiority of supply chain manufacturing production patterns which is of great significance. Therefore, study the scheduling supply chain integration has extensive reality background and important theoretical significance and academic value.In this thesis, on the basis of recapitulate the methods of research of domestic and international supply chain optimization and the field of supply chain collaboration scheduling, using the quantitative method to study the use of supply chain integration model and scheduling methods. Expound the ideas of supply chain modeling and optimization of scheduling. Furthermore, provide specific arithmetic and models dealing the problem of several representative issues.The innovation of this thesis includes:(1) Integrated multi-stages supply chain scheduling model, Consider the integrated model optimal arrangement of multi-link under production and distribution. Take the integrate cost and time as the object of optimization, and set up a multi-objective optimization decision-making model, on this basis, with the supply chain cost increasing situation, created a quadratic programming model of optimal scheduling.(2) We selected the pivoting algorithm as a quick way to solve the non-linear optimization model for scheduling constraints. Pivoting algorithm is more suitable to dealing the problem of large scale model of the scheduling. By solving multi-objective linear programming model and scheduling model of quadratic programming with pivoting algorithm, we found the calculated results is much more superior and can find a optimal solution in a short time.(3) Apply the nested grouping genetic algorithm to large-scale chain integration to solve the problem of scheduling. In the practice, under the situation of supply chain integration, manufacturers and to optimize the process of distribution is not only an assignment problem but also provided with the features of scheduling problem. This brings a lot of calculation problems to solve the model. Here is a nested genetic algorithm, on basis of this, finding the optimal scheduling method to solve the scheduling problem.(4) The WEB integrated scheduling system framework based on Multi-agent consider a supply chain integration based on the manufacture of multi-agent supply network subsystems and functional subsystems framework, in light of the actual demand, bring forward a Bid-based consultation mechanism found on Token. Members of the overall resource scheduling can be rapidly achieved within the use of such mechanism.更多还原