Research On Control Technology Of Two-axis Fast Steering Mirror System Based On High Precision Identification

Fast steering mirror works between the light source or beam receiving end and the target,and is used to adjust and stabilize the optical line of sight in the optical system.Due to factors such as large rotational inertia,shaft system gap and friction disturbance,the common two-axis tracking frame has only tens Hertz of response bandwidth and low positioning accuracy,and cannot overcome the problem of precise beam pointing misalignment caused by high-frequency disturbances in the azimuth and pitch directions.The two-axis fast steering mirror has the advantages of small size,compact structure,sub-millisecond response speed and sub-micro-radian positioning accuracy.and high frequency jitter of the beam in the pitch direction.From the perspective of control technology,it is of great significance to continuously explore and optimize the control technology of the two-axis fast steeting mirror system for improving the system performance indicators.In view of the structural design of the two-axis fast steering mirror system,the first-order mechanical resonance frequency in the X(Y)axis direction is lower than the desired closed-loop bandwidth.At the same time,factors such as the design,processing and assembly of the two-axis fast steering mirror system will inevitably cause the coupling phenomenon between the X-axis and the Y-axis,which reduces the positioning error of the system.In order to improve the fast response capability and positioning accuracy of the two-axis fast steering mirror system,an accurate transfer function model is provided for the design of the single-axis robust controller and the two-axis decoupling controller of the fast steering mirror system based on the highprecision system model identification.In this dissertation,the control technology of two-axis fast steering mirror system based on high-precision identification is studied,and the following results are obtained:(1)The existing physical modeling methods can only obtain the low-frequency spectral characteristics of the fast steering mirror system and ignore the mid-frequency and high-frequency spectral characteristics.Aiming at this problem,this dissertation proposes a piecewise progressive iterative Levy downhill identification algorithm.In this dissertation,the principle,advantages and disadvantages of Levy identification algorithm,Vinagr identification algorithm and Sanko identification algorithm are theoretically analyzed.In view of the shortcomings of the above three identification algorithms,an identification model of piecewise progressive iterative Levy downhill identification algorithm is established.The low frequency band of the model is equivalent to a second-order oscillation link and a first-order inertial link,or two firstorder inertial links or three first-order inertial links,and the middle and high frequency bands are equivalent to several resonance links.The model is divided into several subsystems in the low frequency,medium frequency and high frequency bands,and the iterative Levy downhill identification algorithm is used to gradually identify the transfer function of each subsystem in the low frequency,medium frequency and high frequency bands.Reconstruct the identified sub-systems on the low frequency,medium frequency and high frequency bands to form an accurate and reliable high order system model.The method proposed in this dissertation is compared with the Levy identification algorithm,the Vinagr identification algorithm and the Sanko identification algorithm.The experimental results show that the identification algorithm proposed in this dissertation has the highest identification accuracy for the open and closed-loop frequency response curve of the fast steering mirror system.(2)In the two-axis fast steering mirror system,there are usually factors such as the first-order mechanical resonance frequency in the X(Y)axis direction that is lower than the expected closed-loop bandwidth,the measurement noise of the system,the resonance links in the middle and high frequency bands,and external disturbances,which lead to difficult to improve the bandwidth and robustness indicators of the system.Aiming at this problem,a zero-phase tracking series integral augmented adaptive Kalman filter state feedback control method is proposed.The adaptive Kalman filter is used to estimate the state optimally of the fast steering mirror system,and all the states of the controlled object are constructed.According to the separation principle,the state feedback design method is used to design the feedback parameters of the optimal state estimation.Due to the lack of integral link in the fast steering mirror system,there is a steady-state error in the adaptive Kalman filter state feedback control.Therefore,the integral link is introduced to form the integral augmented adaptive Kalman filter state feedback control to effectively eliminate the steady-state error of the system.In order to reduce the influence of the integral link on the rapidity of the system,zero-phase tracking is further introduced to form zero-phase tracking series integral augmented adaptive Kalman filter state feedback control to improve the system’s fast response capability.The zero-phase tracking series integral augmented adaptive Kalman filter state feedback control proposed in this dissertation is compared with the four control methods of PID,incomplete differential PID,PID+DOB and PID+ARC.The experimental results show that the control method proposed in this dissertation has the smallest step response rise time,the highest closed-loop bandwidth,the smallest error in tracking sinusoidal signals,and the strongest resistance to external random disturbances.(3)Due to the defects of the flexible support itself,the installation error of the actuator and the sensor and other factors,the two-axis fast steering mirror is a coupling system with two inputs and two outputs.The coupling between the two axes leads to the positioning error of the mirror.Aiming at the coupling problem between the X-axis and the Y-axis in the two-axis fast mirror system,a static dual feedforward compensation and dynamic dual neural network adaptive composite decoupling control algorithm is proposed.In this dissertation,the coupling source between the X-axis and the Y-axis composed of DC-coupled components and non-DC-coupled components is analyzed,and the coupled physical model of the two-axis fast steering mirror system is established.The static feedforward compensation decoupling control algorithm is used to offset the DC Coupling components,using dynamic neural network adaptive decoupling control algorithm to offset the non-DC coupling components of the system.The static dual feedforward compensation and dynamic dual neural network adaptive composite decoupling control algorithm proposed in this dissertation is compared with the PID decoupling control algorithm,the dynamic dual neural network adaptive decoupling control algorithm and the static dual feedforward compensation decoupling control algorithm.The experimental results show that the coupling degree and positioning error of the decoupling control algorithm proposed in this dissertation are the smallest.