Research on Fault Diagnosis of Analog Circuit Based on Volterra Theory and Higher-Order Spectrum Analysis

Abstract

To diagnose fault of nonlinear components in analog circuits, a method of fault diagnosis based on Volterra series theory, high-order spectrum and kernel assurance criterion (KAC) analysis was proposed. Firstly, input and output signals of the analog circuit were used to identify a Volterrra model. Secondly, the kernel function was solved by the improved multi-pulse excitation method. The high-order spectrum and its slice method were used to qualitatively characterize the kernel function. Finally, the concept of KAC indicator was introduced to quantitatively compare the characteristics of soft faults in the analog circuit, and the fault degree information was extracted and diagnosed. Taking the Sallen-key filter circuit as the research object, the simulation of the method shows that the high-order spectrum and KAC index can be used to quantify the Volterra system with different fault levels, which can effectively distinguish analog circuit with different soft faults, and have certain practical significance.