Active Fault Tolerant Control of MuPAL-a Using Sliding Modes

This paper proposes a simple adaptive sliding mode observer to estimate the effectiveness level of actuators and uses this information as part of an active fault tolerant controller. These observers create an FDI scheme at a `local' level and the effectiveness estimates are used to drive the online control allocation component in the overall scheme. The approach has been tested on a model of JAXA's MuPAL-α experimental aircraft. The nonlinear simulation results, in fault free and faulty situations, show the efficacy of the scheme. Furthermore, the proposed sliding mode observer has been tested offline using previously collected MuPAL-α flight data and good results are achieved.