Research Of Positioning Control For Piezoelectric Impact Mechanism Based On Neural Networks

Micro-nano technology has been widely applied in many fields of fiber optic docking,semiconductor manufacturing,and medical devices.Piezoelectric actuator has been widely concerned in the micro-nano positioning locations,which the traditional motor,hydraulic and pneumatic equipment positioning accuracy can not meet the requirements.According to the driving principles,piezoelectric actuator can be divided into the ultrasonic motors,inchworm motors,macro-micro motor,and impact drive motors.These kinds of actuators have different advantages and disadvantages.Taking into account the simplicity,rapidity,low cost,theoretically unlimited displacement and the possibility of batch fabrication,piezoelectric stick-slip actuators hold the most promising for accomplishing a compact micro-motion device.A neural network PD controller is designed and used in controlling SIDM.After analyzed different driving theory motors,the piezoelectric stack characteristics are introduced.The mode of vibration is analyzed and the function of vibration mode and resonant frequency is harvested.And the displacement function of SIDM is analyzed when the SIDM is driven by the ramp signal.The valid step is decrease when the SIDM is driven by the ramp signal.So,the trapezoid signal is applied for driving the SIDM.A SIDM prototype is made after analysis the piezoelectric stack characteristic.The working principle is introduced and the stick-slip motion is analyzed.The prototype vibration mode is analyzed by the ANSYS.The resonant frequency is measured.The ADAMS software is utilized in analysis the stick-slip motion.The measured data by ANSYS and ADAMS is applied in setting the experiment condition.In order to realize the positioning control of the motor,the open loop test system and the neural network PD closed loop control system are designed and built.The open-loop test system is used to test the critical stick-slip time of the motor;the closed-loop control system is used in the positioning control of the motor.The closed-loop control system combines the neural network algorithm with the PD controller.The identification process uses the neural network identifier,and the control process adopts the neural network PD controller.The control process includes macro-positioning and micro-positioning control process.The macro positioning control adopts the trapezoidal wave drive signal control.The micro-positioning control process adopts the high level drive signal control,and the two control processes cooperate with each other to realize the positioning research of the motor.From the experimental study,the designed control system can realize the Micro-nano positioning result,and can reach the positioning accuracy of nanometer scale.At the same time,after changing the load quality,the same positioning results can be achieved,which illustrates that the control system constructed in this paper has certain robustness and can meet the requirement of micro-nano positioning under certain working conditions.