Modeling And Inverse Control For Hysteresis In Piezoelectric Actuator Based On Adaptive Filter

Piezoelectric ceramic actuator (PEA) is a new kind of micro-displacement positioningdevice. It has been widely used in micro-electro-mechanical systems, precision positioning,nano-biotechnology, aerospace, instrumentation, and other fields for its high accuracy,small size, low power consumption and fast frequency response, etc. Meanwhile it hasnonlinear properties such as hysteresis and creep. And hysteresis characteristics willseriously affect the accuracy of the nonlinear system. It is important to study the modelingand control of hysteresis characteristics in PEA.Piezoelectric ceramic actuators micro-positioning system experimental platform (PZT)is seen as the research object, modeling and inverse compensation control of hysteresis inPEA based on adaptive recursive filter and adaptive lattice filter are mainly studied.First, modeling algorithm of adaptive recursive filter is explored here. Then the delayoperators are replaced by hysteresis operators (Backlash operators) to compose a noveladaptive structure. And the experiment is carried on the PZT. The modeling effectiveness iscompared with the PEA hysteresis modeling based on adaptive transversal filter. During theadaptive process, LMS algorithm is adopted to adjust the weight values.Then, the adaptive lattice filter is adopted as adaptive recursive filter may be instabledue to the feedback. The modeling algorithm of adaptive lattice filter is explored andgradient adaptive lattice (GAL) filter modeling method is introduced. In order to comparethe modeling effectiveness conveniently, gradient adaptive lattice joint processing (GALJP)method which consists of GAL part and LMS linear part is introduced. To improvemodeling effectiveness, NLMS algorithm is adopted. Then the delay operators are replacedby Backlash operators to verify the modeling effectiveness on the PZT.Afterwards, control algorithm especially adaptive inverse control (AIC) method isintroduced. And hysteresis inverse compensation control is mainly explored. In this paper,cascade inverse model compensation control method and inverse model feedforwardcompensation control method are adopted. Then the experiment is carried on the PZT toverify the control effectiveness.