PREDICTIVE RELIABILITY ASSESSMENT FOR AN AUTOMOTIVE SPEED SENSOR USING HIERARCHICAL TIMED COLOURED PETRI NET

Abstract

For a manufacturer, a good component is one that fails one day after the warranty expires. For a user, a good component means a component that never fails. To predict reliability of a single component produced by a manufacturer, is to attempt to answer an all-important question: when wills the component breakdown? In this paper we will attempt to answer this question by exploring the application of timed hierarchical coloured petri nets to estimate predictive reliability of a regenerative braking system’s electronic component using a stochastic process. The proposed approach is tested on the example of an automotive speed sensor. This approach appears to be efficient for evaluation of stochastic MTBF indicator. Since Monte Carlo simulations are used to provide data for analysis, the present study is not restricted by the Markovian hypothesis.