| The study of effective combination rules in D-S evidence theory when evidences are conflicting is always an interesting topic. This research is concentrated on analyzing the degree of conflict among evidences in Multi-attribute Information Fusion System, proposing a new conflict expression and researching the the D-S evidence theory application in Multi-Attribute Group Descion Making based on Genetic Algorithm. The major works are as follows:(1) Discussed what really constitutes conflict among two beliefs, and defined the new conflict coefficient and distance between the weighted bodies of evidence and proposed the new conflict expression function based on traditional conflict coefficient and evidence distance, this is to say, give the definition and judgement by conflict coefficient and evidence distance, what's more, proposed a global conflict coefficient, computed the optimal weight of the bodies of evidence based on genetic algorithm.(2) Proposed the method of computing the general basic belief assignment based on the triangle fuzzy number. This method firstly used the decision-making system language variables corresponding evaluations triangular fuzzy number represents its alternatives's rating, then, made the triangular fuzzy number of language variables and the planar triangle corresponding corresponding, finally, calculated each triangular fuzzy number in two-dimension plane within and -axes surrounded the trapezoidal area and the area of the triangle intersecting part, and through the unitary im into representative alternatives is optional general basic probability assignment. The simulation results prove that the method can be effectively applied to information fusion.(3) Proposed a fuzzy multi-attribute group decision making model based on general multi-attribute group decision making. Made the effect of conflicting evidences in the information fusion system less by the new improved method in this paper. Simulation result is very good. |
PDF Full Text Request