A Model on Information Fusion Based on Bayesian Network and Game Theory
Yu Ji · Journal of Jiangxi Normal University · 2010
Aiming at the problem of multi-source information fusion,a game fusion model in the form of Bayesian network based on game theory has been propesed.In the study of algorithm for game fusion,according to the requires of game fusion model,the original-node-judgement algorithm have been improved and parzen windows fusion algorithm and integrate them together to form a new Situation Assessment algorithm,and analyse the time complexity in theory.Finally this thesis gets across a prediction example,in which the Bayesian network will be applied to traffic state forecast.In view of traffic flux,speed,traffic density multi-dynamic problems for building suitable Bayesian network model to achieve state traffic projections and to get the forecast simulation results,which compares the original data to be evaluated.