| Traditional inertial actuator for small satellite consists of several reaction wheels.This control method limits the size and weight of the satellite,and increases the difficulty of the attitude control algorithm.The rotor of permanent magnet spherical motor(PMSM)is not mechanically restrained.By assigning the direction of the axis and angular velocity of the rotation,it is possible to provide torque in any direction for controlling the attitude of the satellite.Thus,it can be regarded as a universal reaction wheel and has a wide range of application prospects in aerospace area.The traditional control strategies of spherical motor mainly use the Euler angle as the attitude description quantity,which leads to inevitably singularities in the large angle movement.To conquer the problem,this paper uses the quaternion to describe the spherical motor attitude,and establishes the motion model which consists of dynamic equations and kinetic equations based on the characteristic of quaternion and Euler rotation theorem for rigid body.In addition,the finite element and nonlinear fitting are applied to build the torque model and the back electromotive force(back EMF)model of spherical motor.Since nonlinearity and high coupling exist in the motion system of PMSM,a closed-loop control system is constructed by utilizing sliding mode control algorithm.According to the target of the trajectory tracking,the sliding mode switching function is designed.By selecting the proper reaching law,the sliding mode control law is deduced.And the stability of the system with sliding mode controller is verified.Based on that,a fuzzy controller is introduced to adjust the coefficients of the discontinuous term of the sliding mode controller in real time according to the system state,and to smooth the control signal and optimize the control effect of slide-mode controller.Furthermore,feedback linearization is employed to decouple the motion system of spherical motor,and an extended PD controller is applied to construct a closed-loop control system.Considering that parameters drift,model errors and external disturbances would affect the control behave of feedback linearization controller,a disturbance observer which is based on Butterworth filter is employed to compensate the disturbances.Simulation results show that the two algorithms have good control effect on the spherical motor.DSP and FPGA are used as the control core,and the nine-axis sensor MPU-9250 is used as the position detecting device to build the spherical motor experimental platform. |
