|
Abstract:
|
Electronics is an integral part of almost all industrial systems such as computers that control the systems, and sensors that monitor and control systems. Electronic circuit component characteristics exhibit manufacturing variability just like mechanical parts, and this variability can lead to malfunctions/failures. The repair history for particular kind of equipment may not be able to suggest which specific component(s) will fix the problem and cost could go up if we replace more components than needed. This paper focuses on developing an easy-to-implement and efficient algorithm to do more accurate circuit fault isolation. With the help of electronic circuit simulator, ngspice, we apply adaptive Markov Chain Monte Carlo (AMCMC) to track and update estimates of the component parameter values to diagnose malfunctions in electronic systems. We demonstrate and validate the models using two electronic circuits. The results show that the bad components of electronic circuits can be accurately and efficiently identified.
|
