ON THE USE OF ADAPTIVE UPDATING RULES FOR ACTUATOR AND SENSOR FAULT-DIAGNOSIS

A novel approach is presented for the fault detection and diagnosis (F DD) of faults in actuators and sensors via the use of adaptive updatin g rules. The system considered is linear time-invariant and is subject ed to an unknown input that represents either model uncertainty or unm easurable disturbances. First, fault detection and diagnosis for linea r actuators and sensors is considered, where a fixed observer is used to detect the fault whilst an adaptive diagnositic observer is constru cted to diagnose the fault. Using the augmented error technique from m odel reference adaptive control, an observation error model is formula ted and used to establish an adaptive diagnostic algorithm that produc es an estimate of the gains of actuator and the sensor. An extension t o the fault detection and diagnosis to cover nonlinear actuators is al so made, where a similar augmented error model to that used for linear actuators and sensors is obtained. As a result, a convergent adaptive diagnostic algorithm for estimating the parameters in the nonlinear a ctuators is developed. Two simulated numerical examples are included t o demonstrate the use of the proposed approaches. (C) 1997 Elsevier Sc ience Ltd.