Application And Research Of Fault Diagnosis Based On Bayesian Network Prediction

The bayesian network is a kind of decision analysis tools, appeared with the development of the influence diagram. It provides the knowledge representation, reasoning and learning method in uncertainty environments and can accomplish the tasks such as decision-making, diagnosis, forecasting, classification and so on. Because in the field of fault diagnosis, uncertainty problems are in the majority, the bayesian network is more and more applied in fault diagnosis in recent years.This paper studies the overview of the bayesian network, including the bayesian network express, advantages and features, learning and reasoning and so on, laying the foundation for the use of bayesian networks to solve actual problems and for the establishment of the data structure and dependence. Secondly, introduces the internationally mainstream algorithms about bayesian network at the present and analyzes some kind of algorithms good and bad points. In the production process, we generally get the small data sets, but the mainstream algorithms about bayesian network mostly for the complete or partial large data sets. According to the fact problem, this paper constructs an algorithm thought of bayesian network based on boostrap and K2 algorithm in order to overcome it. The experimentation about asia network indicates this algorithm proposed practicality and availability. Finally, this paper summarizes the recent domestic and foreign power transformer fault diagnosis methods, utilizes the boostrap and K2 algorithm to establish the new power transformer fault diagnosis model, and use it compared with three ratio transformer fault diagnosis method, being beneficial to our analysis, forecast, decision-making and so on.