Research On Information Fusion Method Combined Neutral Network With Evidence Theory On Characteristic Level

Fault diagnosis is a new multi-subject-crossed technique, it has been developed rapidly in last twenty years, and it has, brought huge benefit. Information fusion is a subject formed recently, it has been researched and applied in many fields. But, it is in starting stage in fault diagnosis. There are much available information in fault diagnosis. Only when the available information is used, can the precision and credibility be improved. So fault diagnosis is a process of information. In fault diagnosis of equipments, problems will appear, such as: lots of data to be processed various faults, difficulty of obtaining knowledge, and low ratio of identifying faults. Information fusion on characteristic level is presented to solve the above problems in this paper. Neural networks and evidence theory are combined to diagnose faults, which improves the agility, efficiency and accuracy of the fault diagnosis.Firstly, This text has described information fusion technology characteristic, form structure and concrete processing method, and application of information fusion in fault diagnosis proves the feasibility and validity from the angle of information theory. Based on this, several methods of information fusion fault diagnosis are discussed in detail.1. Information fusion on characteristic level based on neural networks: the topological structure of neural networks and the Learning method are expounded; especially the characteristic of RBF neural network is introduced, and an online algorithm is proposed which can structure the neural network dynamically.2. Information fusion on characteristic level based on D-S evidence theory: the basic principle, formatting rule, reasoning process of evidence theory are introduced in detail. A method of fault diagnosis based on evidence theory and information fusion is proposed and expounded about how torealize the fault diagnosis system.3. Information fusion on characteristic level combined of parallel neural networks and evidence theory: a method which combined parallel neural networks with evidence theory on characteristic level diagnose faults is brought forward. The principle and the concrete steps of this method are introduced at length. At last, the method is proved to have higher diagnosis accuracy and dependability through an instance.