| There exist lots of decision-making problems with hierarchical structure, and different decision makers on different levels have their own goals. The decisions are made in turn from upper level to lower level, and their decision results are influenced by each other. Multi-level programming was developed to solve the problems with hierarchical structure. Supply chain is complicated system with hierarchical or netted structure, and competition and cooperation always coexist in the game of the partners. The purpose of supply chain management is to reduce the total cost and to enhance the competitive power of the supply chain, by coordinating and controlling the flow of materials, information and financing among nodes of the chain. So, the solutions of the supply chain multilevel programming problems should be cooperative ones. Besides, supply chain coordination schemes or collaborative plans are always worked out in situation of information incompletion or uncertainty. Decentralized decision and uncertain circumstance are key characteristics of supply chain management, and they greatly increase difficulties of the researches.After analyzing the features of distributed supply chain management decision making, this dissertation suggests that contractual coordination and collaborative plan are suitable coordination mechanism of decentralized supply chain. The essential objective of the research is to model the multi-echelon supply chain coordination and plan problems and to work out the interactive negotiation methods to solve the models. The dissertation analyses the application of multilevel programming on supply chain problems, and further, the solution characteristics. Based on multilevel programming technique, it models the supply chain coordination and plan problems for 2-echelon and 3- echelon structures. According to distinguish the cooperation strategies and the circumstance certainty or uncertainty, corresponding interactive negotiation methods are designed to obtain the coordinated solutions, based on different optimization strategies.The main contributions of the dissertation are summarized as follows:(1) Supply chain multilevel programming models are built, and significance of cooperation schemes and which the way to get are also analyzed. The existing models that have been built for supply chain, especially the mathematical models, always ignore the decentralization and hierarchy, and scarce researches only consider 2-echelon structures. The multi-level programming technique is applied to model the multi-echelon supply chain coordination and plan problems. Supply chain of 2-echelon, 3- echelon and multi-followers are modeled, and other modeling issues, such as the disposition of restriction expressions, are discussed as well. Besides, it is especially important to get cooperative solution for supply chain multi-level programming models. The dissertation analyzes the classification of the coordinated solutions, and then it suggests that interactive methods through iterative communication on local information are feasible to achieve cooperation.(2) For supply chain contractual coordination issues, a two-step interactive negotiation method aiming to obtain integrative optimization solution is designed. In the situation of information incompletion, the contract parameters are been worked out through iterative communication and repeat game to archive coordination. It is rational for communicators to reserve their local information, and to game tentatively step by step. In the dissertation, the principle of satisfying degree is applied to transform the multi-level programming problem into single-level. Additionally, local information of models such as objective or restriction functions is avoided from absolutely exposing to other partners by constituting objective or variable satisfying functions. For 2-echelon supply chain coordinated by contract, we propose the two-step interactive negotiation methods. The negotiation models are detailed described, and the negotiation processes as well. In the first step, the bi-level programming model is been transformed into single level programming problem on the principle of satisfying degree. Emphatically, the second step negotiation model is build to maximize the sum of the objects' satisfying degrees, based on the trade-off solution obtained in the first step. The designed method is applied to solve a 2-echelon supply chain coordination problem about price-discount contract. Experimental results show that the algorithm has better performance, compared with other optimization algorithms.(3) For supply chain collaborative plan issues, we propose the two-step interactive negotiation method on the principle of cooperative game theory. The existing mathematical models built for distributed supply chain, have the deficiencies of bad-quality solution or requiring much information. The dissertation applies cooperative game theory to collaborative plan problem. For 3-echelon supply chain collaborative plan, the two-step interactive negotiation algorithm is designed. The negotiation models are detailed described, and the negotiation processes as well. Firstly, the tri-level programming model is been transformed into single level programming problem. Emphatically, the second step negotiation model is built on Nash bargaining approach, with the slack restrictions by relaxing the trade-off satisfying degree obtained in first step. Additionally, a fuzzy Genetic Algorithms is presented to obtain the feasible solutions faster. The designed method is applied to solve a 3-echelon supply chain production-distribution problem. Experimental results show that the algorithm has better performance, compared with other optimization algorithms.(4) For supply chain plan problem in fuzzy environments, a two-loop interactive algorithm with parameters adjustment is designed. In practice, supply chain always operates in uncertainty circumstance, and the value of uncertain parameters in distributed hierarchical decision system is higher sensitive to choose. The dissertation establishes the fuzzy chance-constrained multi-level programming model, and transforms it into crisp equivalent with adjustable parameters. To solve the problem, it proposes a two-loop interactive algorithm to get satisfactory solution by iteratively adjusting parameters. In detail, the algorithm includes two interactive procedures: inner-loop and outer-loop. The former is major for the preference of the DM, realized by fuzzy membership functions reflecting goals and decisions attainments; the latter is for the imprecision of parameters, described by possibility degree. Experimental result validates the feasibility of the algorithm.(5) Two-echelon supply chain with multiple followers is studied to build the models and the respective negotiation algorithms. In practical supply chain management, it is common that multiple competitor in a same echelon. According to decision relevancy and cooperation policy, it distinguishes the supply chain with multiple followers into 3 cases as: multiple followers with no decision-connecting, multiple followers with decision-connecting but no cooperation, and multiple followers with decision-connecting and also cooperation. The models and respective negotiation algorithms are separately studies for the three cases.The research results presented above extend the ideas of supply chain modeling. The studies also enrich the content of interactive negotiation principle for decision conditions with information incompletion and game with cooperation and competition coexistence. And it provides new negotiation models and solutions for supply chain oriented Negotiation Support System.
|
PDF Full Text Request