A Fault Prediction Method for Analog Circuits

Abstract

A fault prediction method for an analog circuit, comprising the following steps: performing Monte Carlo analysis on each component of the analog circuit within a fault-free interval and extracting the energy of each frequency band signal; normalizing the extracted frequency band signal energy to obtain a feature vector; training BP Neural network; determine the types of faults that have a tendency to occur; extract the fault prediction feature vector when the component is at the initial value; extract the fault prediction feature vector when the circuit under test is working; calculate the cosine angle distance to characterize the health of the component; calculate the component when it fails The threshold value of the health degree; optimize the selection of the kernel function width factor of the correlation vector machine algorithm; perform fault prediction on the analog circuit. The present invention can be used in both real-time systems and non-real-time systems; it can predict the faults of linear analog circuits and non-linear analog circuits; it can predict the resistance, inductance and capacitance of analog circuits and other main components for failure prediction.

Description

technical field [0001] The invention relates to a fault prediction method of an analog circuit, in particular to a method for establishing a fault prediction model to predict faults of an analog circuit. Background technique [0002] Analog circuits are widely used in household appliances, industrial production lines, automobiles, aerospace and other equipment. The failure of analog circuits will cause performance degradation, functional failure, slow response and other electronic failures of the equipment. Therefore, it is very necessary to evaluate the state of the analog circuit. [0003] The state assessment of analog circuits generally includes fault diagnosis and fault prediction. Among them, fault diagnosis develops rapidly, and the accuracy rate of fault diagnosis can reach about 99% in a large number of research work. The current research results of fault prediction of analog circuits are generally aimed at specific components of analog circuits, rather than the w...

AI Technical Summary

Problems solved by technology

One difficulty with not being able to make predictions for circuits as a whole is that there are few methods that can accurately describe the degradation, i.e., health degradation, of individual components of an analog circuit

At the same time, there are currently few methods to predict the failure of nonlinear analog circuits

Images