Research On Multi-Agent Intelligent Decision Support Systems Based On Theory Of Evidence

With the rapid development of Internet, IDSS is facing new challenge. While providing more information resources for decision-making, Internet provides a fast and easy way to communicate and make it more easily to select attendants in group decision-making. It also brings up many uncertainty of information. How to set up a system to combine the decision opinions is a new problem of group decision-making under Internet. This dissertation introduces theory of evidence as the basic theory to IDSS, discusses the frame and decision procedure of multi-agent IDSS based on theory of evidence, regards the theory of evidence as the tool to realize the multi-agent cooperative decision that aims at the problems that may exist its application to the corporative reasoning of multi-agent IDSS. And researches the combination of evidence based on theory of evidence. This dissertation combines the theory of evidence with the theory of rough sets and Grey to acquire the basic probability assignment and to solve the tough problem how to obtain the evidence in the theory of evidence. At the same time, it also study the decision-making method based on theory of evidence to make the decision scientifically. The detailed research contains:(1) Researches the method of evidence's acquisition in theory of evidence. Firstly, through further research of the Rough Sets theory, we refine the granule of the classification, and then change the decision attribute of the decision table. Base on these and the relationship between Rough Sets theory and Evidence theory, we compute the evidence's basic belief assignment through classification, and thus we realize the acquisition of evidence. Then, we combine the theory of evidence with the theory of Grey provide another method how to obtain the evidence in the theory of evidence. So we can solve the tough problem how to acquire the evidence in the theory of evidence.(2) Researches the rule of combination in theory of evidence. Aiming at the theory of evidence's limitations that the rule of combination requires independent evidence and it can't deal with conflicting evidence, can't distinguish the importance of evidence, researches the accordingly improved method and improves the combined effect of D-S formula.(3) Researches decision methods based on theory of evidence. It discusses adecision method through solving the basic probability number based on the focal analysis, relates the resolve of basic probability number with process of decision making to rationalize decision method which based on theory of evidence.(4) Researches the framework of multi-agent IDSS based on theory of evidence. This dissertation researches the model system of multi-agent IDSS's framework based on the theory of evidence. Aiming at the key question of MAS how to organize and collaborate among multi-agent, this dissertation poses the multi-agent united solving model that based on the frame of conceptional tree. Besides, it also reduces the solving extension of conceptional tree's layers and discusses the possible question from three aspects of frame of discernment, relativity of evidence and combination formula according to the applied characters of the theory of evidence under the multi-agent, and gives the corresponding methods of resolution.