A method for fault diagnosis of analog circuit based on rough set

Abstract

Based on rough set reduction artificial immune system, a new method for fault diagnosis of analog circuit is proposed. The proposed method uses wavelet to analysis the output voltage as the fault examples. Then the examples are reduced through attributes reduction to obtain a smaller set of all examples. And the new samples are trained to get the optimal cluster center of each fault. Finally, the fault component is located by comparing the distance between the test samples and the optimal cluster centers. The simulation result shows that the proposed method has high accuracy in diagnosis of tolerance analog circuits, and higher speed than pure artificial immune system.