Supply Chain Network Design And Analysis With Risk Control

Supply chain network design involves typical two-stage decisions,namely,the facility location decision and the distribution pattern decision.The first-stage location decision has to be made in the case of some important information uncertainty,such as customer demands and operating costs.The second-stage distribution pattern should be determined accordingly to the location decision and obtained information,such as the realizations of the customer demands and operating costs.Because the location decision has to be made before the realization of uncertain information is observed,the location decision is inherently associated with risk.In order to evaluate the risk in the two-stage decision process owning to the uncertainty of information,the concept of Value-at-Risk(VaR)is introduced for the two-stage supply chain network planning problem,and further,a new accurate solution approach is proposed.The major study contents of the thesis are summarized as the following three parts.Firstly,a new solution approach to VaR-based two-stage fuzzy location problem.It is proved that the fuzzy two-stage programming model involving regular fuzzy numbers can be transformed into an equivalent deterministic one-stage programming model,thus the infinite-dimensional optimization problem can be converted to a classical finite-dimensional problem,by which the computation complexity of solving the two-stage model is significantly decreased and meanwhile exact solutions can be obtained.Secondly,VaR-based fuzzy two-stage model for supply chain network planning and its accurate solution approach.A fuzzy two-stage programming model with VaR criterion is proposed for controlling the investment risk in supply chain network planning,considering the uncertainty of customer demands and transportation costs.The objective is to minimize the VaR of investment loss.In this regard,the proposed model is solved via converting into an equivalent deterministic one-stage model.Numerical examples with different sizes illustrate the effectiveness of the proposed model and solution approach.Thirdly,analysis for the location decisions and the loss mation.Some new results for practical management are deduced by analyzing the location decision and the loss distribution.The value of fuzzy information is clarified by comparing the two-stage decision based on fuzzy information and using the expected value instead of fuzzy information.The proposed two-stage programming model and accurate solution approach are general enough to deal with the supply chain network planning in various industries under uncertain environment.