Hysteresis Nonlinear Identification Of Piezoelectric Stack Actuators And Active Micro-vibration Control Method

With the rapid development of micro-nano manufacturing technology,the nanoscale technology in the field of micro-machining has become mature.In the process of ultra-precision machining and testing of micro-workpieces,the requirement of production and testing environment become more stringent.The micro-vibration between the environment and the related equipment is one of the factors that seriously affect the precision of machining and testing.How to isolate these micro-vibrations safely and reliably from micro-nano machining and testing equipment has become one of the key problems to be solved in the field of ultra-precision equipment and micronano equipment.The active vibration control platform based on piezoelectric stack actuator has been paid more and more attention,in order to overcome the shortcoming of passive vibration control.In this dissertation,the hysteresis nonlinear identification and active control of micro-vibration isolation platform based on piezoelectric stack actuator are studied,a discrete Bouc-Wen model and a discrete asymmetric Bouc-Wen model are established to describe the hysteresis nonlinear behavior of piezoelectric stack actuators.A parameter identification method of Bouc-Wen model based on improved particle swarm optimization is proposed,an affine projection adaptive filtering algorithm is introduced to study the active micro-vibration control based on piezoelectric stack actuators.The main work is as follows:(1)An improved particle swarm optimization algorithm is proposed by combining with Jaya algorithm,which does not need to adjust any parameters of the algorithm itself,based on the analysis of the implementation and performance of particle swarm optimization algorithm.The simulation results of three test cases show that the improved particle swarm optimization algorithm has better convergence ability to the global optimal solution,and can realize efficient optimization solution and model parameter identification.(2)The Bouc-Wen model and the asymmetric Bouc-Wen model are developed to describe the hysteresis nonlinear characteristics of the piezoelectric stack actuator,and the discrete Bouc-Wen model and the discrete asymmetric Bouc-Wen model are derived for the discrete control system requirements.The improved particle swarm algorithm is used to identify the parameters of the two discrete Bouc-Wen models,and the proposed discrete asymmetric Bouc-Wen model describes the hysteresis nonlinearity with higher accuracy compared with the discrete Bouc-Wen model.(3)The adaptive filtered-X affine projection algorithm(FXAPA)is studied,and the corresponding derivation formulas and analysis are given.Combining the idea of variable step-size and convex combination theory,a variable step-size FXAPA algorithm and a convex combination of variable step-size FXAPA algorithm are proposed,and the corresponding computational complexity are given.The FXAPA algorithm,the variable step-size FXAPA algorithm and the convex combination of variable step size FXAPA algorithm are simulated and compared,and the effectiveness and superiority of the proposed convex combination of variable step-size FXAPA algorithm in active micro-vibration control are verified.(4)A three-degree-of-freedom micro-vibration active isolation platform based on MATLAB x PC real-time control environment,piezoelectric stack actuator,capacitive displacement sensor,power amplifier,and signal conditioner is constructed.The FXAPA algorithm,the variable-step FXAPA algorithm and the convex combination of variable step-size FXAPA algorithm are used for the experimental vibration active control study,and the corresponding real-time vibration control effect is obtained,based on this active vibration isolation platform.The experimental results show that the proposed variable-step FXAPA algorithm and convex combination of variable step-size FXAPA algorithm have better vibration suppression performance than the FXAPA algorithm,and the convex combination of variable step-size FXAPA algorithm has more superiority.