Discrete‐time polynomial systems approach to combined fault monitoring and control

Abstract

AbstractThe combined fault monitoring and control problem is discussed, using an H2 optimization framework. The main aim is to investigate the frequency‐domain and pole‐zero properties of the combined control and fault estimation compensator. The solution of the LQG stochastic optimal multivariable control and fault estimation problem is obtained using a polynomial matrix systems approach. The combined control and fault estimation problem involves a performance index including the usual regulating error and control weighting terms and also fault estimation error terms. The system model includes input disturbances, reference, coloured noise signals and actuator/sensor fault signals. The proposed numerical algorithm has common polynomial terms, so that the computation of the fault estimates and control law, is not significantly more difficult than the calculation of the control law alone. Copyright © 2003 John Wiley & Sons, Ltd.