Fault Detection And Isolation Using Unknown Input Observers With Structured Residual Generation

Abstract

Detection and isolation of faults are important tasks for improving the reliability of process industries in order to enhance their products quality. This paper investigates the detection and isolation of faults using structured residuals. Actuator and sensor faults are considered. Residuals are generated using a bank of unknown input observers (UIO). Three-tank benchmark system was used as a prototype of many process industries. Simulation results show the effectiveness of the studied method.