Failure diagnosis and nonlinear observer. Application to a hydraulic process (Q1867787)

Fault diagnosis is an issue of primary importance in modern process automation since it provides the basis for safe and reliable operation of complex engineering systems. Failure diagnosis consists of providing information on the time and location of faults occurring in supervised systems, and is naturally divided into fault detection and fault isolation. This paper addresses fault detection and isolation for nonlinear systems. The authors' approach is based on techniques of nonlinear disturbance decoupling and on the use of nonlinear observers. The authors give sufficient conditions and a design procedure for synthesis of residual generators. Then they characterize a class of nonlinear systems for which a high gain observer can be designed. Finally, a hydraulic process example is used to illustrate detection and isolation of three errors via the authors' methods.