| Supply chain (SC) is a value chain composed of raw materials suppliers, plants, Distribution Centers (DCs), customers, in which information, capital, logistics flow over, so the SC deals with the problems involved of supplier selection, product quan-tity control, location of manufacture plant and DC, material requirements planning etc. In the management of real life supply chain, many factors exert some kind of uncertainty, and often the uncertainty presents as fuzziness.This thesis gives an expected value supply chain model and a VaR supply chain model, in which the former solves the problem under a meaning of expected value while the later deals with the minimum risk value problem with a preselect credi-bility level. We develop an approximation method to turn the original two infinite dimensional supply chain problems into approximate finite dimensional problems respectively, and discuss the convergence property of the approximation objective function value. Furthermore, we design an AA-based particle swarm optimization (PSO) to solve the proposed two-stage fuzzy SC models and some numerical exam-ples are given to demonstrate the effectiveness of the designed algorithm.The major new results about the thesis include the following several aspects:(ⅰ) We study and present several definitions on the feasible region of two-stage fuzzy programming, and we give some properties about feasible region.(ⅱ) Based on different modeling criterions, we investigate two new class of two-stage fuzzy supply chain models:the expected value supply chain model and the VaR supply chain model.(ⅲ) We present the computational methods of the expected value objective function and the VaR objective function in the two kinds of fuzzy supply chain models, and prove the convergence of the objective function of the approximating SC model to that of the original one.(ⅳ) We design a hybrid PSO algorithm, and give some numerical examples to illustrate the effectiveness of the designed algorithm. |
PDF Full Text Request