Fault detection method based on recursive Fisher information matrix

This paper proposes a control system fault detection method based on information geometry theory. By analyzing the correlation between the Posterior Cramer Rao Bound(PCRB) of the system and the Fisher Information Matrix(FIM), a recursive model of the Information Flow Matric(FIM) was established to integrate closed-loop control information and kinematic equations of the system. A fixed wing unmanned aerial vehicle(UAV) flight control simulation system is built based on the digital platform, and the sensitivity of the changing trajectory of the recursive FIM determinant value to the abnormal dynamics of the system was verified under typical actuator faults. This method provides a new basis for constructing fault detection thresholds with physical significance. A new research approach has been developed for the design and optimization of high reliability UAV flight control systems.