Intelligent Control of Satellite Formation Flying
Small satellites flying in formation present a more efficient and affordable way of achieving the same or better performance than a large satellite because of low cost, high density of functionality and a short development cycle. A key technology for achieving mission objectives is the attitude and orbit control system. The overall objective of this dissertation research focuses on developing advanced control strategies and fault tolerant control for satellite formation flying.It is necessary to design and operate the satellite formation flying system to reduce fuel consumption and improve control accuracy. This is a very challenging task due to the nonlinear nature of satellite formation dynamics and the risk of thrusters’ failures and sensors’ faults in the absence of hardware redundancy. A class of nonlinear leader-follower satellite formation flying systems subject to uncertain thrusters’ and sensors’ faults and external J2 disturbances has been studied applying fault detection and identification and second order sliding mode control methodologies. New fault detection and identification and fault tolerant control algorithms were compared with model based fault detection and identification and fault tolerant control algorithms in presence of large initial errors, timevarying external disturbances, and parameter uncertainties. The faults considered were modeled as constant or ramp faults. Numerical results demonstrated the effectiveness of the proposed active fault tolerant control under actuators’ and sensors’ faults. It has been shown that the proposed second order sliding mode control scheme can guarantee local asymptotic stability after system faults. Simulation results confirmed that the suggested control methodologies yield high formation keeping precision and effectiveness for leaderfollower formation flying systems. The tracking errors of the proposed second order sliding mode control, adaptive fuzzy sliding mode control, chattering free sliding mode control and classic sliding mode control resulting from the thruster faults are within 2 m, 4 m, 10 m and 1 m, respectively. The fuel consumption of the proposed second order sliding mode control was the least.It is also necessary to design a fault tolerant satellite attitude control system to reduce fuel consumption and improve control performance accuracy. The proposed fault tolerant attitude control algorithms were based on first order and higher order sliding mode control theory as well as fuzzy logic systems to achieve real time autonomous fault tolerant control. These algorithms were applied to attitude synchronization in both leader-follower formation flying and decentralized formation flying. Attitude synchronization during formation flying was examined considering actuator dynamics while decentralized attitude ynchronization was studied using graph theory with quaternion kinematics. The proposed fault tolerant control algorithm was compared with the existing satellite attitude system controllers in the literature and it was found that the proposed algorithm resulted in three axis attitude stabilization within 0.041◦ in all axes for the fault cases. The reaction wheels’ Coulomb friction, saturations, noise, dead-zones, bias fault and external disturbances are considered. Finally, a nonlinear adaptive fuzzy sliding mode controller was tested using embedded nanosatellite hardware on a frictionless spherical air bearing system. The test results showed attitude errors of 0.8◦ using the proposed controller while a proportional integral derivative controller resulted in 5◦ attitude errors.