Assignment Problem In Double Uncertain Environments And Its Algorithm Research

Assignment problem is one of the classic decision problems in the operational research and widely exists in real life. For traditional assignment problem already has a relatively mature research, these problems can be used by the methods, such as hungarian method and implicit enumeration method. In the process of making the operation and management decisions, the assignment problem of uncertain conditions becomes the focal point of many scholars' present research. The uncertain environment in complex reality no longer just has random factors and fuzzy factors but mix them together, so it creates the assignment problem under a dual uncertain environment.Assignment problem under the double uncertain environment is mainly divided into random fuzzy and fuzzy random two cases study. The specific priority is as follows:First of all, on the basic of uncertain programming theory, a kind of dual uncertain assignment problem in practical application is studied. By the analysis of the random fuzzy assignment problem that based on the resource constraints, the expected value model is set up, and a equivalent algorithm under certain conditions to solve the model is designed. Numerical solution is presented, which demonstrates the rationality of the model. Secondly, by studying other models of assignment problem under random fuzzy environment the chance constrained programming model and related opportunity programming model are set up respectively. In order to solve the model,the hybrid intelligent algorithm is designed, which combines the random fuzzy simulation with genetic algorithm. And the effectiveness of the algorithm is verified through a specific numerical example. Finally, the assignment problem under random fuzzy environment is studied. On the basic of uncertain planning theory, modeled on the modeling process of random fuzzy environment, several different models for the assignment problem under random fuzzy environment are built. The rationality of the model is proved through a specific numerical example.