An on-line fault detection scheme for systems with known nonlinear state-space models

In this paper, a simple scheme for the detection of faults and changes in operating point for processes with well-known nonlinear state-space model will be described. The emphasis will be in the solution of practical problems. A linearized Kalman filter, the statistical t-Test and some decision rules are utilized in the scheme, which is validated in the case of a practical application to a laboratory turbo-generator.