Research On Fault Tolerant Control Algorithms For LEO Drag-free Satellite With Actuator Failures

Since the 1960 s, the development of space technology has brought great changes to the basic scientific research. The weightlessness of space has created a unique environment for some basic physics experiments. As the absolute free falling movement does not exist in space, the object will be subjected to a variety of interference force(such as the resistance of the interstellar medium, the solar radiation pressure and the cosmic ray radiation pressure,etc), the drag-free satellite is designed to offset the disturbance of non gravitational force by corresponding thrusters, which can create more close to the absolute free fall for scientific experiments.Since there are many uncertain factors that will be affected by the drag-free satellite flying in space, propellers as the main actuator, which would be failed inevitably. Unknown actuator failures can seriously affect the performance of the drag-free control system, in order to make the system run safely and stably in space, a controller has to be able to accommodate those failures and compensate the effect caused by them automatically once the failures taking place. Several approaches have been achieved to control the systems with actuator failures, see for instance, sliding mode control-based scheme, model predictive control, adaptive control and neural network control, etc. This paper will investigate the adaptive fault tolerant control problem for the LEO drag-free satellite control system with unknown actuator failures.Because of the complex structure of the drag-free satellite, and it will be subject to uncertainty when flying in space. So the nonlinear kinematic model of the drag-free satellite is established, and the state space expression of the system is obtained, then the system is divided into three independent subsystems, we will design the adaptive controller for each subsystem. In consideration of the actuators with unknown failure, we use backstepping control theory and Lyapunov stability theory to design the parameter estimation and control rate, which can achieve the required control objectives. The designed controller can effectively compensate for the unknown actuator failures, so that the system has better anti-jamming performance and robust stability, the simulation results verify the effectiveness of the controller.