Uncertain Environment Emergency Logistics Location And Transportation Optimization

The selection of facilities location and the transportation optimization is closely linked. It is a direct impact on the performance of rescue work. This paper will study the facilities location combined and vehicular traffic routing problem under the uncertain environment, to realize the optimization of the entire emergency logistics system. This method has the theoretical significance and research value of practical application in degree.First, this paper set out the joint optimization tools for the uncertain characteristics of emergency logistics, namely stochastic programming and bi-level programming theory, to solve the selection of facilities location and the transportation optimization under the uncertain environment. Then, according to the decision making of emergency logistics, a bi-level programming model of stochastic chance-constrained is presented. Finally, this paper put forward a method to translate the multi-zoned into the single-planning, so it is a basis for solving stochastic bi-level programming model of LRP.This paper research the model of basic location and its relationship. Starting from the particularity of emergency logistics, the set covering model under uncertain environment is set up to solve facility location.According to the characteristics of natural disasters and emergency events, this paper constructs a Bayesian networks for situation assessment of the emergency events, EM algorithm is used to determine the probability distribution of two Bayesian network parameters-demand and vehicle running time.Papers focus on specific issues of this study, based on the above-mentioned stochastic programming and bi-level programming theory, the LRP stochastic bi-level programming model is set up. In this model, the upper research the vehicular traffic routing problem, and the lower research the selection of facilities location under emergency to realize the facilities point at least as far as possible the constraints of maximum coverage. The process of genetic algorithm is designed for the established model. Finally,the application of the model and its algorithm are valid with a practical example.