| Analog circuits are used in all walks of life widely and become an important element in the modern science and technology, it is harder for fault diagnosis of analog circuit with smaller volume and more integration because of the analog circuits with nonlinear, continuous and tolerance properties. The existing methods are still not satisfactory and many similarities for fault feature extraction of analog circuit, they are not characterized thorough and comprehensive for fault features with the large-scale circuits of the development and the status quo and node reduction,based on this reality, the fractal theory is used for fault diagnosis of analog circuits.Among them, the fractal dimension can measure and identify fault state of circuit effectively.Based on the fractal theory, researches mainly focus on two aspects of fractal dimension and Fractal Fault Diagnosis method.In view of the existing methods exist more similarity, respectively by fractal box dimension and information dimension of analog circuit fault for feature extraction, i give based on the box dimension, the information dimension of the features of the circuit and the corresponding obtain the fractal parameter of the circuit.For the box dimension and information dimension is not from different angles and at different levels reflect analog circuit fault state, I use multi fractal theory for fault diagnosis of analog circuits and use multifractal spectroscopy for fault feature extraction and fault diagnosis method.This article gives three circuit features based on the fractal theory, master of the acquisition circuit of the fractal parameter method and corresponding fault diagnosis methods, and the typical circuit simulation. The results show that the fault diagnosis method is feasible. It also provides a method for fault diagnosis of analog circuits and has certain theoretical significance and practical value. |