【作者】 张国权；

【导师】 李文立；

【作者基本信息】 大连理工大学 ， 管理科学与工程， 2010， 博士

【摘要】 随着供应链内外不确定性的增加,仅提高供应链的绩效已无法适应供应链的发展,供应链不确定性管理由此受到了广泛的关注。目前关于不确定应对策略与技术方面的研究,一般只限于特定的对象,或者特定对象的某一方面,缺少多方位的不确定考量,如绝大多数的研究集中于需求不确定下的供应链的协调方法与策略,此外许多关于供应链建模的研究都基于理想化的假定。本研究将从供应不确定、需求不确定和供需不确定引发的制造不确定这三个方面研究以制造商为核心的供应链多策略响应模型。本研究目的在于提出一个较为完整的多策略响应系统模型,即在供应链不确定识别及其特征分析的基础上,建立供需不确定的生产响应策略模型、供应不确定的采购响应策略模型和供应商的优选方法、需求不确定(依赖于价格和购买行为)的销售响应策略模型。具体地,本文开展了以下几个部分的研究。(1)建立了供应价格不确定和需求服从正态分布的生产响应模型,提出了求解该模型的综合智能算法。首先,针对一个现实的一般化的以制造商为核心、原料价格模糊且最终市场需求服从正态分布的供应链网络,建立从原料采购到最终产品销售的一系列供应链计划模型；其次,针对目前求解算法存在的局限性,提出了分散进化算法,并将其与模糊优化和随机机会约束规划相结合而构成了综合智能算法,以求解提出的模型；最后,将提出的方法与其他方法作比较,从实验角度验证了模型的合理性和求解方法的有效性,并对参数进行了敏感性分析。研究发现：随着供应市场变化的加剧,综合考虑供需不确定可以动态优化供应链的供需结构,为供应链带来利润和客户服务方面的整体提高。此外,提出的求解方法能够很好的解决复杂的、现实的、大规模问题,可以为类似问题提供解决方法。(2)提出了供应违约不确定下基于机制设计理论的采购合约响应模型,设计了二次订货策略下最优的长期与短期合约及各条件下合约的选择。将供应不确定中最为典型的供应中断风险为出发点,首先设计了供应中断风险信息完全或不完全条件下基于机制设计理论的两部制的显示合约,重点研究二次订货策略下违约信息不完全或信息完全的长期合约与短期合约设计问题,以及二次订货时各条件下最优的合约设计与合约选择；最后将以上需求确定条件下的研究扩展到需求不确定情况,并且考虑两阶段信息更新条件下最优合约的设计。本研究发现：在违约信息作为供应商的私人信息的情况下,作为Stackelberg主导方的制造商可以通过基于机制理论的两部制合约设计来缓解潜在供应中断风险,且在信息完全条件下制造商提取了整个渠道的利润,在信息不完全条件下,制造商提取整个渠道的部分利润,且可靠度最低的供应商的利润始终为零。供应不确定下二次订货策略是重要的响应策略,且长期合约不一定较短期合约对制造商更有利,决策者需要根据一定的外部市场参数来决定合约类型。(3)供应不确定条件下,提出了基于信息粒度熵的供应商优选方法。选择适合的供应商是应对供应不确定的另一策略。首先通过专家的模糊语言评价得到三角模糊评价矩阵,在解模糊化和K均值聚类的基础上,利用待优选对象的评价数据本身所具有的信息量确定权重,然后使用TOPSIS对方案进行排序,以便最终选优。最后将该方法用于不确定环境下供应商的优选过程,并从7个方面将信息粒度熵方法与AHP进行了比较和分析,实例说明了此方法的合理性和有效性。研究发现：信息粒度熵方法利用待优选对象的评价数据本身所具有的信息量,通过挖掘数据内部所蕴含的信息确定权重,使得评价结果更为合理。(4)针对需求依赖于价格和购买行为的问题,提出了最优的团购销售响应模型和双渠道销售响应模型。在需求实现过程中,需求不仅与价格有关而且与需求行为特征有关,因此本研究利用在参与群体购买过程中形成的羊群行为,通过采纳团购销售策略最大化利润。基于模型的求解获得团购销售时最优的价格和数量包,并将即时购(个体购买)与团购相比较,得到不同市场参数下的最优销售策略。此外,将提出的模型应用于单一企业多产品情况,分别得到团购和个体购买各种策略组合下,最优的价格和数量。在此基础上,分析替代性产品之间的竞争以及纳什博弈均衡解。最后分析各种策略下社会福利的变化与社会有效性,并与最优社会效益进行比较。本研究发现：1)羊群效应较等待成本因子对价格和利润的影响更大；2)在双渠道策略下,团购策略和个体购买策略呈现互补性与替代性；3)在竞争市场下,各种竞争策略,如团购vs.团购和团购vs.个体购买,存在均衡的价格和数量,并且应用超模博弈理论证明竞争存在唯一的纯纳什均衡。本文重点从数学建模与优化的角度,研究了供应和需求不确定的以制造商为核心的供应链多策略响应问题,以提高供应链反应和应对能力,为有效地管理各种潜在不确定提供可行的途径,丰富和完善了供应链不确定性管理理论。更多还原

【Abstract】 With the increase in supply chain uncertainty, supply chain management can not only concentrate on supply chain performance improvement, but also need to pay attention to supply chain risk or uncertainty management. Risks may come in a form of economic instability, environmental concerns, volatile fuel costs, strike, work stoppages, supply shortages, and quality concern. The current coping strategies and methodologies on supply chain uncertainty are generally limited to a specific objective. For example, the majority of the supply chain research focuses on coordination methods and managing supply chain under demand uncertainty. Most of them lack a comprehensive deliberation and are based on idealized assumptions. This dissertation will study multi-dimensional response models of supply chain from the perspective of manufacturer. The three response models address the supply uncertainty, demand uncertainty, and production uncertainty caused by supply and demand uncertainty.The purpose of this study is to propose several multi-dimensional response models. By characterizing the uncertainties in the supply chain, this research establishes a number of response strategies to tackle various scenarios under uncertain supply and demand. The details are as follows:(1) The research established the multi-echelon production planning model for manufacturer under supply price and demand uncertainty. We proposed a scatter evolutionary algorithm that integrates fuzzy stochastic chance-constrained programming to solve the model. First for a generalized supply chain network with fuzzy raw material prices and normally distributed market demand. a set of planning models were established from raw material procurement to final product sales. Due to its complexity, an integrated solution framework which combines scatter evolutionary algorithm, fuzzy programming and stochastic chance-constrained programming are combined to jointly take up the issuel. Finally, a computational study is conducted to evaluate the model. Numerical results using the proposed algorithm confirm the advantage of the integrated planning approach, and the impacts of uncertainty in demand, material price, and other parameters on the performance of the supply chain are studied through sensitivity analysis.. We found that taking into account supply and demand uncertainty jointly can optimize the supply and demand structure to increase supply chain profits and improve the overall customer service. In addition, the proposed method can provide good solution for complex, realistic, large-scale problems.(2) Under the partial or full supply uncertainty information, this research applies mechanism design theory (reverse game theory) to address the manufacturer’s contract coping strategy for long-term and short-term contracts selection. This research considers supply disruption as the most common and detrimental risk in supply uncertainty. We first study the contracts under the risk of supply disruptions with perfect and imperfect information based on the two-part price mechanism and mechanism design theory. We identify the design parameters that maximize profit for the manufacturer under the default supply uncertainty, and determine the amount the manufacturer with incomplete information is willing to pay to obtain the full information. Next we examine the long-term and short-term contract design strategy if secondary ordering is allowed with incomplete and complete default information. We also optimize contract design and contract selection under different conditions. Finally, we extend above results to the conditions with uncertain demand, and consider the two-stage information updating in the optimal contracts design. This study found that given the disruption information as suppliers’private information, the Stackelberg dominant manufacture can alleviate the supply disruption risk using the proposed two-part contracts based on the mechanism design theory. Under the complete information condition, the manufacturer could extract the entire channel profit. The manufacturer can extract partial channel profit under incomplete information condition, and the profit of the supplier with high disruption probability is zero. We also found the secondary ordering can be used as another strategy to respond to the supply uncertainty for the manufacturer. Under certain conditions, long-term contracts are not more favorable than short-term contracts to manufacturers. The decision makers thus need to determine the contract type according to external parameters.(3) In order to cope with the supply disruption risk, the author developed an Information Granulation Entropy-based model for supplier selection. In the proposed model, experts input fuzzy language to form an evaluation matrix. After defuziffying the matrix, the K-means clustering method is applied to discretize the matrix. An innovative information granulation entropy approach, based on information science theory and data mining technique, is developed to determine the weights of criteria. Finally, the TOPSIS closeness rating method is applied to derive the priorities of the alternatives. To demonstrate the validity of the proposed method, the author illustrates the model with a real-world application faced by a large company for selecting a supplier, and compared and analyzed the method with AHP in 7 aspects. The proposed evaluation framework is especially beneficial when dealing with large-scale problems with diverse criteria and/or alternatives. The study found that this method extract the information of evaluation data to weight, which validates the result of evaluation.(4) For the endogenous product demand that depends on the price and dynamic group behavior, we proposed an optimal Group Buying strategy and dual-channel marketing response model. For price-sensitive demand, we take into account the impact of group opinions on customers’purchasing decision. The herd behavior often leads to distorted demand, which in turn affects pricing under various market conditions. We proposed retailers various optimal pricing and packet-size determination models. Along with instant buying (individual buying), we contrast and determine the optimal selling strategy for different market parameters. Under the MIX strategy GB is complementary and substitutable for IB. We extended these models to multi-product business cases, and obtained the optimal price and quantity under various combinations of Group Buying and Instant Buying. On this basis, competition between merchants of substitutable products is analyzed. It is found that the competition has a Nash Equilibrium Solution. We also analyzed the payoff of both competitors under the conditions of the various parameters, and thus obtained competitive equilibrium solution. We found that the impact of group pressure to a customer is generally much greater than the waiting costs in the Group Buying strategy. Finally social welfare and social effectiveness were analyzed for the various strategies.This dissertation focuses on the multi-dimensional responses of the manufacturer as the core of the supply chain, using the mathematical modeling and optimization techniques under supply and demand uncertainty to improve the responsiveness, and provide a viable way to effectively manage a variety of uncertainty and risks, and enrich the supply chain uncertainty management theory.更多还原