Font Size: a A A

The Study Of Analog Circuit Fault Diagnosis System Based On Information Fusion

Posted on:2017-05-23Degree:MasterType:Thesis
Country:ChinaCandidate:L J LiuFull Text:PDF
GTID:2348330491962667Subject:Control engineering
Abstract/Summary:PDF Full Text Request
With the development of electronic technology, the degree of complexity of the modern equipment is increasing, and the later maintenance of electronic circuit is particularly important. In the practical applications, due to the factors like nonlinearity, susceptibleness to outside interference, the failure of analog circuits occupy a large proportion in the total equipment failure. Therefore, the in-depth study of fault diagnosis of analog circuits has greatly important significance.Firstly, the development of the fault diagnosis of analog circuit was reviewed. At the same time, it introduces the feature extraction methods, the neural network and the Multi-information fusion technology. Aiming at the shortcomings of the traditional information fusion system, it combines variety of feature extraction methods, the neural network, with D-S evidence theory of the information fusion, and builds the fault diagnosis system based on three levels of information fusion technology.Secondly, the input of BP neural network is the circuit characteristic information, which is extracted by the fusion of PCA and wavelet packet analysis. The diagnostic results of BP neural network are made as the preliminary results, and provide the information of the evidences for the final diagnosis of D-S evidence theory. So, it actualizes the connection among the feature/feature-making, the feature/decision-making and decision/decision-making.Moreover, it makes the application of the fault diagnosis system based on information fusion technology to fault diagnosis of analog circuits, and describes the process of fault diagnosis with specific circuit simulation. The simulation results of circuit output the preliminary diagnosis results of BP neural network and the final diagnosis of D-S evidence theory fusion. The simulation results show the method has higher diagnostic accuracy and faster speed.Finally, it actualizes the application of three levels of information fusion technology of fault diagnosis system to the fault diagnosis of radar's analog circuit, and builds the structure. It describes the operation of the software of the fault diagnosis to the radar by specific instances, and shows the design is of humanization and intelligence.
Keywords/Search Tags:Information Fusion, Fault Diagnosis, Analog Circuit, Neural Network
PDF Full Text Request
Related items