Fault Diagnise Of Satellite Attitute Control System And The HIL Simulation

During the long-term operation of the satellite,which is exposed to harsh space environment,lots of failures and malfunctions may occur,also a variety of new technologies and new concepts put forward a high demand for satellite safety and reliability,which makes the fault detection and diagnosis(FDD)more and more important.In the failures of satellite,the satellite attitude control system faults occupy a large proportion,but because the satellite attitude control system is difficult to fully model,only the mathematical simulation can not verify the validity of the algorithm in the physical system.Therefore,this paper establishes a hardware-in-the-loop(HIL)simulation system of satellite attitude control system based on the three-axis air-bearing based platform,and uses the mathematical and the HIL simulation to verify the TSEKF and ATSUKF in the satellite attitude control system fault diagnosis.The main contents of the thesis are as follows:A HIL simulation system of satellite attitude control system based on the three-axis air-bearing based platform is constructed to verify the fault diagnosis algorithm.This paper introduces the working principle and design of the three-axis air-bearing based platform,including the hardware composition and software environment.The moment of inertia of the platform is measured and can be used for subsequent experiments.In the HIL simulation of the attitude control system,the actuator/sensor faults are simulated,including the drift in the fiber optic gyroscope,and additive or multiplicative faults in the reaction wheels.The influence of fault on attitude control system is analyzed.Two-Stage Kalman Filter(TSEKF)is derived and applied to the fault diagnosis of actuator and sensor faults in satellite attitude control system.In the mathematical simulation,the effectiveness of TSEKF is verified while considering the reaction wheels additive or multiplicative faults,as well as the faults in attitude angle or attitude angular velocity measurement.Due to the linearization,there is a deviation in the fault estimation of sensor faults using TSEKF.In the HIL simulation,TSEKF is applied to estimate the friction torque of reaction wheel,and the relationship between the friction torque and the reaction wheel speed is obtained.After compensating the friction torque of reaction wheel,TSEKF is used to estimate the actuator/sensor faults in HIL simulation,which is verified to be effective.In order to solve the problem of fault estimation in strong nonlinear systems,UKF is introduced into Two-Stage Kalman Filter,and Two-Stage Unscented Kalman Filter(TSUKF)is proposed.The mathematical simulation results show that the proposed algorithm is better than TSEKF while estimating the gyroscope and star sensor faults.In order to estimate fault when the system information are incomplete,this paper uses the adaptive matrix to adjust the noise covariance matrix and propose the ATSUKF fault estimation algorithm,which effectively avoid the decrease of the performance of the algorithm when the system noise covariance is unknown.Finally,the HIL simulation of TSUKF and ATSUKF algorithm is carried out,which verifies the feasibility of the algorithm in physical system.