Abstract: For highagility and highprecision attitude maneuver of remote sensing satellites, the control moment gyroscope (CMG) is expected to supply more accurate torque, which means the disturbance torque acting on its gimbal axis should be minimized to reduce the fluctuation of the gimbal speed. The torque generated by the wheel imbalance is analyzed, and the gimbal system model is built. To improve the gimbal speed stability, an iterative learning control (ILC) module is inserted into the PID gimbal control system. The proposed control system is evaluated via system simulation, and the results demonstrate that the ILC algorithm can minimize the fluctuating torque effectively, which significantly improves the precision of the CMG output torque.

Key words: CMG, iterative learning, pulsation torque minimization