The Method Study For Fault Diagnosis In Analog Circuits Based On Information Fusion Technique And Wavelet Analysis

In the paper,fault features are abstracted by wavelet analysis,then neural network and Information fusion are used to local fault competences.The main contributions in the paper are as follows:1.We designed a typical BP neural network fault diagnosis system for analog circuits and it can detect whatever fault in theory.However,in order to construct a high performance system,it is very important to extract significant features to compactly represent the circuit behavior.2.A feature extraction method of "energy-fault" diagnosis based on the wavelet transform is presented,which overcomes the disadvantages that the traditional diagnosis approaches need topology structure of analog circuit. Simulation results show that this new method is available for fault diagnosis for analog circuit.3.In order to solve the Problems of insufficient test data and low diagnosis accuracy,a fault location approach for analog circuits is developed based on intelligent information fusion technology including data fusion and neural network.Input node current and output node Voltage are abstracted by wavelet analysis,then feature fusion are used to local fault competences.The proposed approach makes full use of the fault information and has the capability to diagnose multiple and soft faults of analog circuits with satisfactory accuracy.