Isolation and Handling of Sensor Faults

Abstract

This chapter considers the problem of sensor FDI and FTC for nonlinear systems subject to input constraints. The key idea of the presented method is to exploit model-based sensor redundancy through state observer design. An output feedback control design using high-gain observers is first presented; and then an FDI scheme is presented, which comprises a bank of high-gain observers. Residuals are defined as the discrepancies between these state estimates and their predicted values based on previous estimates. A fault is identified when all the residuals breach their thresholds except for the one generated without using the measurements provided by the faulty sensor. Upon FDI, the state estimate generated using measurements from the remaining healthy sensors is used to preserve practical stability of the closed-loop system. The implementation of the sensor FDI and fault-handling framework subject to uncertainty and measurement noise is illustrated using a chemical reactor example.