Research On The Hysteresis Modeling And Control For PZT-Actuated Two Dimension Micro Platform

The micro-platforms which are actuated by PZT actuators have many advantages of fast response,repeatable operation,high resolution etc.Therefore,they have been widely applied in the precision positioning situations as core components,such as microscopic measurement and manufacturing,ultra-precision tracking and controlling,and become the technical strength to push the implementation of cutting-edge technology.However,there exists serious hysteresis nonlinearity in PZT-actuated micro platform between the input voltage and output displacement due to its structure and material properties.It can degrade the operation or positioning accuracy and become the important element of instability.Given this,the research of this paper mainly focuses on the exploration and research of two aspects: establish a much more accurate model to describe hysteresis nonlinearity and design an effective control strategy to eliminate or compensate the hysteresis nonlinearity.In the current study,a brief presentation about actuators,displacement transmission mechanisms,hysteresis modeling and compensation strategies have been made sufficiently.On the basis of this,a discrete mathematical description of Krasnosel'skii-Pokrovkii(KP)hysteresis model is proposed,and the weights are identified online by Adaptive Linear Neural Network(ALNN).Furthermore,in order to improve the modeling accuracy,the KP element operator has been modified by adding a function of describing the creep characteristic.The advantage of KP model in this study is the online description of hysteresis nonlinearity.At last of this section,with the PZT-actuated two dimension micro platform as controlled object,a validation experiment of KP model is carried out at the environment of rapid control prototyping(RCP)simulation system.Furthermore,a KP-based iterative learning controller(ILC)is designed to compensate the hysteresis nonlinearity.The iterative update principle is based on the output of KP model.According the asymmetrical characteristic of hysteresis nonlinearity,different iterative update principles are selected for the increase and decrease stages of voltage.For the improvement of control performance,an integer order PID compound controller of KP-based ILC compensation is developed to eliminate the hysteresis nonlinearity further.Experiments are carried out at three typical reference displacements to validate the feasibility and effectiveness of ILC compensation and integer order PID compound control.The results demonstrate that hysteresis nonlinearity can be compensated effectively,and the integer order PID compound control can improve the tracking precision.Moreover,hysteresis nonlinearity presents unlocal memory characteristics,that is,the current output displacement not only dependents on the current input voltage,but also dependents on the history of the input voltage and displacement.While,the fractional order operator owns similar property,it has the advantage of memorizing the function information over a period of time.Therefore,a fractional order PID(FOPID)compound controller of KP-based ILC compensation is proposed to degrade the bad influence of hysteresis nonlinearity.The experiments results show that the FOPID compound controller has a faster adjustment speed and a minimum tracking error when compared with the integer order PID compound control.The main drawback of FOPID compound control is that there are lots of parameters need to be identified.To overcome this problem,a fractional order adaptive controller is designed.The adaptive law makes full use of the output history of KP model,and the involved system parameters can be obtained through online adjusting.The control scheme is verified experimentally under different forms and frequency of reference displacements.The results suggest that the fractional order adaptive controller can eliminate the hysteresis nonlinearity and improve the tracking control precision o more effectively.