On Degradation Fault Diagnosis Method For Analog Circuits

With the development of the electronics industry, the development of electronic components is towards digitization, integration and miniaturization. The complexity of electronic products and equipment are increasing. Analog circuits still play an important role and occupy an important position due to some of their irreplaceable in the entire supporting system. Theoretical analysis and practical application show that, analog circuit faults are more common, and analog circuit test cost much more than digital circuit. Analog circuit faults are usually caused by the component parameter drift, this type of fault is not easy to detect, but is the hidden trouble affecting the system reliability and safety. If we can timely detect the hidden danger, and make targeted maintenance and repair, so as to ensure the safety and reliability of the system, but also greatly save detection time and repair costs. Therefore, research of analog circuit fault diagnosis technology has very strong actual demand and realistic significance.Using computer simulation platform and focusing on the practical application, this paper focuses on degradation fault diagnosis technology for analog circuits, mainly on the following question.(1) To diagnosis degraded fault for linear analog circuits, through constructing a comparison circuit, a new fault dictionary based on voltage sensitivity is established; then fault localization and fault parameters identification are realized for multiple faults case. The fault components can be localized and identified by analyzing the voltages of available test-points.(2) Hilbert-Huang Transform is applied to the analog circuit fault diagnosis and a fault feature extraction method is proposed based on Hilbert-Huang Transform, the fault components can be localized combined with BP neural network. This method can not only diagnosis single fault, but also diagnosis multiple faults, just needs one accessible test point, not only applicable to linear analog circuits but also applicable to nonlinear analog circuits and diagnosis accuracy is ideal. (3) An analog circuit fault parameter estimation method is proposed based on Hilbert-Huang Transform to achieve single fault parameter estimation. We can establish the parameter estimation equation based on the relation equation between the various component parameters and corresponding output voltage signal Hilbert marginal spectrums. After location the fault component, substitute the Hilbert marginal spectrum energy of the output signal into the parameter estimation equation and solving, the fault parameter can be obtained. The approach is applicable for nonlinear circuits as well as linear ones. The parameter estimation accuracy is high and just needs one accessible test point, is suitable for practical engineering applications.For each part of the study, simulation or actual circuit experiments are done to verify the feasibility and effectiveness of the proposed method. The experimental results show that the proposed various methods achieve established goals, and meet the needs of practical engineering applications to a certain extent.