In this paper, a method for real-time fault detection and isolation is presented, which has been tested successfully for unmanned aerial vehicles. In this method, two parity vectors are constructed to achieve the fault detection and isolation. In addition, the maximum likelihood estimate is used to estimate and correct the fault vector. Simulation results show that not only can the faulty system be detected accurately in time by this method, but also the fault can be repaired efficiently. Thus, the fault tolerance of the integrated system is improved remarkably by this method.

No abstract

No abstract