Teaching Model-based Fault Detection and Isolation using a Virtual Laboratory Environment

Abstract

Fault detection and isolation (FDI) systems play a key role to provide efficiency, reliability and safety in today’s industrial processes. The teaching of FDI systems is facilitated if it is carried out not only with theoretical lectures but also with practical experiences. This paper proposes a virtual laboratory environment (VLE) to carry out online practical experiences with FDI systems for a benchmark process. Thanks to this VLE, students can set up faults in sensors, actuators or in the process itself, program model-based FDI algorithms and test FDI system performance. The use of this environment is illustrated by testing the performance of FDI systems for the quadruple-tank process (4TP) under different fault scenarios. Finally, the procedure of using this proposal for practical experience with two model-based FDI design methods is shown.