Research Of Hysteresis Compensation Control Method For Piezoelectric Ceramic Micro-positioning Platform

With the speedy development of Nanotechnology, piezoelectric ceramic micro-positioning platforms as a core component of precision manufacturing equipment attract more and more attention. However, the hysteresis nonlinearity occurring inherently in piezoelectric ceramic micro-positioning platforms has a significant negative impact on the performances of positioning accuracy, response speed; Therefore the further application and development of the piezoelectric ceramic micro-positioning platforms are hindered. The research is concentrated on how to eliminate the bad influence of hysteresis nonlinearity in the piezoelectric ceramic micro-positioning platforms. Hysteresis nonlinearity modeling and advanced controller design are determined as the two main research priorities after studying the hysteretic compensation control theories and methods deeply.In this paper, an introduction about the structural performance and working principle is presented firstly. Thereafter, a conclusion of hysteresis nonlinear modeling and control methods is drawn. In order to describe the hysteresis nonlinearity of piezoelectric ceramic micro-positioning platform accurately, in the current research, Dynamic Prandtl-Ishlinskii(DPI) hysteresis model with dynamic rate related properties is proposed on the basis of classical Prandtl-Ishlinskii(PI). The density parameters of DPI hysteresis model are identified by bat optimization algorithm based on the measured data of the piezoelectric ceramic micro-positioning platform. Whereas there exists disadvantage in the process of the identification that the bat optimization algorithm is liable to trap into local optimum resulting in the precision of DPI hysteresis model cannot satisfied the requirements. To address this problem, Wavelet neural network(WNN) algorithm is selected instead of the bat optimization algorithm to identify the parameters of DPI hysteresis model. Simulations results show that the DPI hysteresis model identified by WNN algorithm has a lower modeling error and can predict the hysteresis characteristic well.According to the relationship of stop operator and play operator in PI hysteresis model, a DPI inverse hysteresis model is established on the basis of DPI hysteresis model. The density parameters of DPI inverse hysteresis model are identified by WNN algorithm and the DPI inverse hysteresis model also has a good dynamic characteristics. Then, a feedforward controller is designed based on DPI inverse hysteresis model and three different signal are adopted to test the effectiveness of the feedforward control. Simulations carried out in this paper suggest that the hysteresis nonlinearity can be mitigated effectively and the piezoelectric ceramic micro-positioning platform control system presents a pseudo-linearity characteristic.It is well known that feedforward control cannot reduce the model errors and has a bad anti-interference performance, therefore, a hybrid control strategy based on DPI inverse hysteresis model and PID control are investigated for a higher positioning precision. The PID parameters are adjusted in real-time by WNN algorithm. Simulation results validate the feasibility and effectiveness of the hybrid control.For the purposed of improving the system robust performance and increasing the positioning accuracy, WNN adaptive robust control based on backstepping control method is developed. The critical point of this control method is that WNN algorithm is selected to approximation system unknown function, and then design the corresponding adaptive and control laws. In addition to this, the global stability of the piezoelectric ceramic micro-positioning platform system is proved by the Lyapunov function. Simulation results demonstrate that comparing with the hybrid control strategy mentioned above, the WNN adaptive robust control can more effectively improve the positioning precision of piezoelectric ceramic micro-positioning platform.