Aero Starter Generator Fault-diagnosis Based On Volterra Series Model

The novel aero starter generator fault diagnosis method based on Volterra model was proposed in this thesis.Firstly, the non-linearity dynamic differential model of the aero-starter-generator in motor state is established. Then the simulated analysis of four stage of the aero-starter-generator in starting state is proposedIn the third chapter, the basic definition of Volterra model, which used to detect and diagnosis faults in what kinds of non-linear dynamic systems, is proposed. Then the development of Volterra model to detect and diagnosis faults is introduced.In the forth chapter, the Volterra modelof aero starter generator was used to detect and diagnosis faults. By using the steepest descent rule, the estimated second-truncated Volterra kernel parameters can be obtained after the object function converged. Based on the above results, the geometric assignment of kernel parameter vector corresponding to particular fault or normal operation, which is lies in a fixed and fault-specified direction in the parameter space, was used to fault diagnosis and Fault Isolation.After many experiments, it proved that this method was feasible, which provided a new way and method to detect and diagnosis faults of aero starter generator fault diagnosis...