Two-level Programming Model Based On Satisfaction

In this thesis,a new satisfaction function is proposed by using the limit tolerance value and ideal value of the upper and lower level objective function in the bilevel programming model,which has fewer variables and is easier to be solved.The satisfaction function is applied to the following two situations.Firstly,we construct a new partial cooperation model with the lower satisfaction function as the cooperation coe cient in order to solve the ill-posed bilevel programming problem.By solving the numerical example,we find that the overall satisfaction is better if the minimum satisfaction level is changed.This partial cooperation model under the satisfaction degree is feasible.Secondly,in the resource allocation bilevel programming model,the degree of environmental pollution is taken as a fuzzy variable,the production cost and the product price are taken as random variables,and an opportunity constraint bilevel programming model is constructed at a certain confidence level.Then the equivalent theorem is used to transform the model.Further on an interactive algorithm based on satisfaction function is proposed to transform the bilevel programming model into a single level programming model and solve it.Finally,sensitivity analysis is made based on confidence level and upper satisfaction level.