Robust Adaptive Control Method For Micro-vibration Of Multi-degree Of Freedom Structure Based On Piezoelectric Stack

With the continuous development of space industry,the active micro-vibration control of the spacecraft itself,the equipment and other space devices have become one of the important research hotspots in the current aerospace field.As the core content of active vibration control,the research of active vibration control method covers almost all control theories and techniques.Considering the possible nonlinear and time-varying characteristics of the complex space devices,adaptive control method has become an important branch in active micro-vibration control of space devices.And some positive research results have been obtained in this field.However,due to the difference in research background,application,structure and other aspects,the research methods and implementation techniques adopted are also different.The relevant technical method has only been realized preliminarily and has their own application limitations,so it is urgent to explore the relevant theory and technique further.Based on the project named "Robust Adaptive Control Methods and Techniques for Micro-vibration of Space Devices" supported by National Natural Science Foundation of China,this dissertation focuses on "improving the robustness of the adaptive control methods for micro-vibration control system." An active microvibration control experimental system based on piezoelectric stack with multi-degree of freedom was constructed.The theoretical analyses and experimental verifications of multiple input and multiple output robust parameter adaptive algorithm,multiple input and multiple output closed-loop and online identification algorithms,multiple input and multiple output robust adaptive micro-vibration control algorithms were completed.The main research work of this dissertation is as follows:(1)For the active micro-vibration control of the multi-degree of freedom structure,a co-simulation platform was built through the joint operation of ADAMS and MATLAB.Through simulations and analyses with the co-simulation platform,an active micro-vibration vibration control structure with multi-degree of freedom based on piezoelectric stack actuators was manufactured.Then,an active micro-vibration control experimental system was constructed to verify the feasibility and reliability of the relevant methods and techniques.(2)For the lack of robust consideration of the existing parameter adaptive algorithms in multiple input and multiple output active micro-vibration control system,a robust multiple input and multiple output least squares algorithm combined dead zone and normalization was proposed,and the corresponding mathematical analysis and implementation were given also.The result of co-simulations and experiments showed that the improved least squares algorithm has better robustness than the original edition.(3)Considering the model parameter identification of the secondary path of active micro-vibration control system with multi-degree of freedom,this dissertation studied and derived closed-loop output error algorithm,filtered closed-loop output error algorithm and extended closed-loop output error algorithm.At the same time,an online model correction algorithm for secondary path of active micro-vibration control system was presented.The result of co-simulations and experiments showed that the three multiple input and multiple output closed-loop error identification algorithms were effective.Also,the on-line model correction algorithm for secondary path of active micro-vibration control system could significantly improve the convergence speed of the adaptive control algorithm and relax the accuracy requirements on system identification of secondary path.(4)This dissertation studied the multiple input and multiple output robust adaptive algorithm based on internal model principle.Then,the derivations and implementations of feedforward and feedback robust adaptive control algorithm based on Qparameterization were given.Combining the advantages of feedforward control and feedback control in active micro-vibration control,the multiple input and multiple output hybrid robust adaptive control algorithm based on Q-parameterization was deduced and its convergence performance was analyzed.Adapting the least squares algorithm combined dead zone and normalization as parameter adaptive algorithm in feedforward,feedback and hybrid active micro-vibration control strategies based on Qparameterization respectively,the improved control algorithms were obtained.The effectiveness and robustness of the proposed algorithms were verified by cosimulations and experiments in active micro-vibration control.(5)For the problem of frequency mismatch in multiple input and multiple output active micro-vibration control under multiple frequency narrow-band disturbance,the existing solution is using parallel adaptive control algorithms.However,this method may cause computational increasing.Based on the idea of hybrid adaptive control,a hybrid adaptive active micro-vibration control algorithm was proposed in this dissertation for frequency mismatch.Co-simulations and experiments showed that the proposed algorithm can achieve better micro-vibration suppression than the conventional parallel adaptive algorithm in various disturbances.