Degradation Trend Prediction of Linear Regulator Based on SVR Under Nuclear Radiation Stress

Abstract

Due to the development of science and technology, many electronic products still need a long time to degrade and fail under the condition of accelerated life test, especially under the harsh test conditions such as nuclear radiation, and it brings great challenges to research on the reliability of electronic products. In order to obtain the performance index of electronic product degradation failure, this paper proposes to use support vector regression(SVR) method to predict the performance degradation index of AP1117E series linear voltage stabilizer under nuclear radiation stress, and use the degradation data obtained from the test and the predicted degradation data to complete the reliability evaluation of the device. The prediction method proposed in this paper is used in the actual reliability assessment engineering project, and it has played a certain suggestive role for future reliability assessment work.