Research On Fault Diagnosis Method And Application On Electronic System Based On Bayesian Network

With the continuous development of microelectronics technology and computer technology, the function of missiles, satellites and other large equipment are gradually perfect and powerful. And their internal structure is becoming more and more complex. Well, in the modern large-scale equipment, electronic system tends to occupy a core position, and its failure often leads to serious consequences. Under the condition that the external test node of the electronic system is getting less and less and the complexity of the electronic system has multiple fault, associated fault and so on,it will lead to the uncertainty problem, making the fault diagnosis more difficult. At this time, the traditional fault diagnosis methods have been ineffective. Therefore, it’s very necessary to study the suitable fault diagnosis method for electronic system.This project comes from a pre research project " system level fault diagnosis technology research on electronic equipment ". It plans for the characteristics of electronic systems and focuses on electronic system fault diagnosis method based on Bayesian network to solve the key technical problems like fault detection, fault location and fault isolation. And the validity of the method is verified by a fault simulation system which is based on a certain type of equipment. In the project, the Bayesian network modeling and reasoning methods are mainly studied, and the fault diagnosis experiment is carried out on the communication subsystem which is made up of some functional modules. Compared to the traditional method s, bayesian network method has unparalleled natural advantages. It can be through structural learning to implement the "black box" system model and use parametric learning to achieve the final fault detection and location. On the basis of that, this paper designs a fault diagnosis simulation platform based on a control system for the prototype and also designs fault injection platform to simulate the real fault injection. Besides, this paper builds the fault collection platform including the excitation data card to obtain fault sample data and develops the fault diagnosis of PC software platform. It finally achieved through the control mode of fault injection, fault collection and fault detection and positioning.Finally, in the fault diagnosis system, the thesis use the method proposed in this paper to do the experiment of modeling and diagnosis. Then it selects an electronic system fault simulation platform as the application object to do the experiment of fault diagnosis modeling, and eventually the experimental results and electronic systems in the presence of uncertain problems are in-depth analysis, verifying the validity of the method presented in this paper. In addition, the research results of this paper can be applied to the basic level maintenance support testing, fault diagnosis and fault repair of all kinds of large-scale electronic equipment. It has a good guiding significance for the fault diagnosis, detection, location and repair of electronic system.