摘要

The important precision control of the interferometer suffers from various faults in the complex space environment. However, the studies for the fault-tolerant control (FTC) in the interferometer are few. This paper first presents an active FTC scheme for the interferometer system in the presence of micro-gravity, micro-vibration and thermal disturbances in space. Kinds of unanticipated faults are quickly detected by the residue detection. Neural network is used to estimate the unanticipated faults online. The sliding mode control with a reaching law is designed for the normal control with a fixed upper bound and for the FTC with an adaptive upper bound adjusted by the neural network to recover the system performance under the faults. Stability analysis is performed by the Lyapunov theorem. Numerical simulations are used to verify the significant improved performance of the proposed FTC scheme.