Research on Fault Prediction Method of Analog Circuits Based on DCQGA-SVR

摘要

Aiming at the shortcomings of analog circuit fault prediction technology, the DCQGA-SVR analog circuit fault prediction method is proposed. Firstly, the failure mechanism of electrolytic capacitor and power MOSFET in analog circuit is analyzed. The fault parameters that can be used to reflect circuit fault, map electrolytic capacitor and power MOSFET are researched. The average output power is extracted as the characteristic parameter of circuit fault. Secondly, the double-chain quantum genetic algorithm is used to optimize the support vector machine regression algorithm to predict the fault feature parameters, and then the fault prediction model is established. Finally, the Boost circuit is used as the simulation test circuit, and compared with the traditional QGA-SVR algorithm and LS-SVM algorithm. The results show that the proposed prediction method has small mean square error and can effectively predict the fault.