Fault Diagnosis of Aero Engine Fuel Regulation System Based on Extended Kalman Filter

The fuel regulating system is used for the automatic supply fuel flow and the adjustment of aircraft flight status. Fault diagnosis of the system is challenging due to its high incidence of fault, variety fault forms, concealment of fault characteristics. The high testing cost and high safety risk caused by failure. This paper presents a fault diagnosis method based on extended Kalman filter to reduce the estimation error introduced by local model linearization and improve the accuracy of fault model. The AMESim simulation model is established, the fault mode is studied, and the typical fault mechanism is analyzed. The linear analysis tool of AMESim is used to solve the state space equation. The nonlinear system model is linearized by the extended Kalman filter, and the internal parameters of the system are observed to estimate the state of the model and determine the system's fault state. To verify the feasibility and effectiveness of the proposed fault diagnosis strategy, different types of fault excitation are added to the model, and the simulation and experiment are compared. An aero-engine fuel regulator fault diagnosis platform based on Kalman filter algorithm was designed. The fault diagnosis rate and accuracy of the algorithm platform meet the design requirements.