경로모델과 확률과정모델을 이용한 전자부품 잔여수명 예측

Abstract

The nuclear power plant is conducting periodic detailed inspections, and in the current state, only the pass/fail degree of the parts and circuits in the electronic card is judged. The instrumentation control card plays an important role in protecting and controlling nuclear power plants. Therefore, it is required to develop and construct a system that can predict and diagnose not only the current state of the nuclear power plant instrumentation and control card but also the progress failure that may occur during the operation of the power plant and perform pre-maintenance. The scope of this study is to determine the degree of degradation of 6 types of components (Photocoupler, BJT, SCR, Electrolytic Capacitor, MOSFET, Digital IC) that cause a lot of failure and predict the remaining lifespan using PHM (Prognostics and Health Management) technique. do. To analyze the degradation model, two major models are used: the Degradation Path Model and the Stochastic Process Model. A model that describes the mechanism of degradation is called a degradation path model. It refers to a model that shows changes over time in the characteristic parameters of parts and products related to failure. The parameters of the degradation model are divided into a deterministic coefficient model and a random coefficient model according to a constant or a random variable. The stochastic process model is widely used in various fields such as finance, meteorology, and aviation. Small particles floating on a liquid were introduced to mathematically model the random molecular motion. In order to compare the analysis results according to the degradation model, the photocoupler compared the degradation path model and the stochastic process model. Among the stochastic process models, data parts suitable for the gamma process were analyzed with the gamma process, and the photocoupler was analyzed with the Weibull process applying the gamma process because the gamma process was not suitable. As a result of the analysis, the electronic component with the longest lifespan was the MOSFET, and the component with the shortest lifespan was the photocoupler with the most failure.