| Aircraft modelling based on Research Civil Aircraft Model (RCAM) is first introduced in this work. The actuator fault isolation problem for the aircraft is then investigated. An adaptive observers (AO) and a sliding mode observers (SMO), were chosen for fault diagnosis purposes. An adaptive scheme based on a bank of AOs is designed to detect and isolate constant actuator faults. Theoretical results are obtained to ensure that the proposed schemes can work well for actuator fault diagnosis under certain assumptions. Simulation results on RCAM show that the adaptive observer based technique can detect and isolate constant actuator faults successfully. However, the adaptive scheme can not deal with non-constant actuator faults (which may also occur in civil aircrafts). It also can not estimate the actuator faults which may be crucial for fault accommodation purpose. As an addictive, in order to deal with non-constant actuator faults and to estimate the faults, an SMO based scheme is proposed for fault detection, isolation and estimation. This method can be used for any types of faults. Under certain assumptions, it is proved the SMO based scheme can detect, isolate, and estimate various actuator faults. A number of simulations studies are performed on RCAM model for various abrupt and incipient actuator faults. The results show that both abrupt and incipient faults can be detected and isolated successfully, which is consistent with the analysis. Moreover, both abrupt and incipient faults can be estimated very accurately. Although fault accommodation is not addressed in this work, it is important to note that accurate estimation of faults can lead to a more solid fault diagnosis decisions and can provide useful information for fault accommodation purpose. |
