Fault Reconstruction For Spacecraft Attitude Control System Based On Sliding Mode Observer

With the increasing of human space activities,the number of on-orbit spacecraft are increasing,the spacecraft's state of health has been widely concerned by scholars at home and abroad.The development of fault diagnosis technology for spacecraft plays an important role in the successful completion of the flying mission.Taking the spacecraft attitude control system as the research background,this dissertation makes a study on actuator fault reconstruction and sensor fault reconstruction based on a series of technologies,such as sliding mode observer,neural network,adaptive control etc.For the actuator fault reconstruction problem of spacecraft attitude control system,the Back-Propagation neural network adaptive sliding mode observer(BPASMO)and Radial Basis Function neural network adaptive sliding mode observer(RBFASMO)are designed based on the linear model and nonlinear model of spacecraft attitude control system respectively.Firstly,a BPASMO is designed to overcome the adverse effects caused by the parameter uncertainties,external disturbance and the condition of unknown upper bound fault.The system state estimation errors can be proved to asymptotically convergence to zero.The actuator fault can be reconstructed based on the designed BPASMO.Then,considering the operating condition that the yaw axis information is incompletely measurable,a RBFASMO is designed based on the nonlinear model of spacecraft attitude control system.Design the coordinate transformation matrix,the system model is divided into two subsystems which the unknown input and fault only contains in the output subsystem.Under the application of equivalent output injection concept,the proposed observer was used for actuator fault reconstruction.Finally,simulation results are provided to illustrate the effectiveness of the proposed schemes.For the problem that the estimated error of the state which cannot be measured is asymptotic convergence in the standard sliding mode observer,an actuator fault reconstruction scheme based on the terminal sliding mode observer is designed using the finite time convergence property of terminal sliding mode.Firstly,considering the loss of effectiveness fault and stuck fault,the spacecraft attitude control system's linear model with actuator faults is introduced.Then,the fractional power is introduced into the design of nonsigular terminal sliding mode observer.The Lyapunov function was given as the judgment condition to enable the non-output state estimation errors finite time convergence.Accoding to the output of terminal sliding mode obsever,the proposed observer was used for actuator fault reconstruction.Finally,Simulation results are provided to illustrate the effectiveness of the method.For the sensor fault reconstruction problem of spacecraft attitude control system,a fault reconstruction scheme based on adaptive sliding mode observer is designed.Firstly,considering the adverse effects caused by the system unknown inputs,non-singular coordinate transformation is introduced to divide the system into two subsystems where the subsystem 1 only includes the effects of system unknown inputs and the subsystem 2 one only includes the effects of system sensor faults.Then,considering the unknown upper bound of the unknown inputs and sensor faults in the subsystems,the adaptive sliding mode observers are designed for the subsystems.In view of the subsystem contains only the sensor faults,the original system is transformed into an augmented dimensional nonlinear description system,and the sensor fault reconstruction problem is converted into system state estimation problem,the parameters of observers are determained by solving linear matrix inequality(LMI).In the subsystem 1,the effects of unknown inputs can be dealed by the discontinuous control input in adaptive sliding mode observer.Finally,Simulation results are provided to illustrate the effectiveness of the method.