Research On Composite Control Strategy Based On High Precision Identification

Recently, rapid development of target detection, navigation, fire control systems, aerospace and space optical communication call for increasingly high performance of photoelectric tracking system. Compound axis has attracted tremendous attention, and its main advantages are that the tracking accuracy and response speed can be greatly improved by combing main system and subsystems under the premise of normal working condition. However, in practical system, the control accuracy improve with the higher requirement of transfer function identification, whether the transfer function model is accurate or not will limit control performance directly. Conventional control algorithm is difficult to balance the relationship between tracking accuracy and response speed, miss distance lag caused by detector seriously affects the tracking accuracy of the system, and los jitter caused by vibration platform reduces the extraction accuracy. Therefore, to solve the problems, this thesis give a deep research on system identification, servo control system design and los jitter analysis, and put forward effective methods. Then compound axis tracking servo system based on dual detector was established and studied to demonstrate the feasible of control algorithm. The main work studied in this thesis is arranged as follows:(1)System identificationFor the problem in transfer function identification, a method based on improved differential evolution algorithm was proposed. The proposed method improve the performance both in the convergence speed and optimization efficiency by changing parameters such as initialization parameters, mutation operators, and crossover factor. Comparing with QPSO, DE, JADE algorithm and classical algorithms, such as: Levy and the direct and indirect algorithm in time domain, MDE algorithm shows better performance in identification results.(2)Servo control technologyFor the main tracking control system, three closed loop control system based on current, speed and position loop was designed. A control strategy based on ADRC is proposed and applied to position loop to achieve high tracking precision and fast response, an extended state observer is introduced in position loop to estimate the unknown dynamics and uncertainties, and then compensate the disturbance in real time. Tracking differentiator is introduced to deal with miss distance delay. Taking into consideration that the frame frequency of detector is restricted much lower than sampling frequency, the compensation results were further predicted, thus effectively improving the tracking performance. To solve the time delay problem in subsystem, a method based on tracking differentiator was proposed and studied, respectively, which improve the dynamic performance. In order to improve the tracking accuracy in subsystem, the control method based on disturbance rejection control is proposed and studied, simulation results show that the algorithm is superior to the conventional control algorithm.In order to verify the correctness and validity of theoretical and simulation results. Composite axis tracking system based on dual axis tracking detector was built and studied, respectively. Numerical experiments in tracking moving target are provided to demonstrate the remarkable performance of the proposed methods in terms of tracking accuracy and response speed when compared with traditional PI control algorithm.(3)Los jitter analysisFor the impact of los jitter on image quality, a new algorithm of the analysis of vibration blurred images based on the image gray-scale is proposed. The simulation results of blurred image caused by los jitter is obtained according to the simulation model. Los jitter device has been established to obtain distorted image. Comparison results between simulation and experiments demonstrate the correctness of image blur algorithm, which provide a solid foundation for the platform stabilization device.