Time-frequency characteristics optimal control of fast steering mirror for image motion compensation

In order to further optimize time and frequency domain characteristics of fast steering mirror (FSM) for image motion compensation and thus to increase the definition of picture taken by moving camera, a control system based on linear extended state observer(LESO), zero phase error tracking control(ZPETC) and Kalman filter was designed. First, mathematical model of FSM driven by voice coil motor(VCM) was established. Next, principles of LESO, ZPETC and Kalman filter were clarified. Finally, experimental research for controlled object was made. Step response curves demonstrate that the settling time reaches 2.5 ms. Besides, compared to proportional-integral-differential(PID) control system, relative error of output converges to zero faster when it is less than 1%. Bode diagram shows that the bandwidth attains 369 Hz, which is more than 1.5 times that of PID control system. Meanwhile, phase lag reduces nearly a half. When disturbance is added, average relative error of the designed system reaches 0.028%, reduced by 75% compared to that of PID control system. Pictures taken by camera illustrate that this control method makes both subjective visual effect and objective evaluation parameters of image improve further.