Abstract:

This paper is concerned with the detection of actuator faults in nonlinear systems. With the theoretical development of sophisticated fault detection and isolation (FDI) approaches, the implementation of FDI algorithms is, however, often constrained by the available computational power. On the other side, in many practical applications, in order to avoid extra hardware and investment, FDI algorithms often share processor with control algorithms. Such a background motivates the study in this paper, namely, whether part of the controller information can be utilized to simplify the FDI algorithm and thus promote the application of FDI technique in industry. In the context of inverse dynamics controller a positive answer is provided by the investigation in this paper