A Robust Position/Force Control System Design of 6DOF Stewart Platform for Nanoscale Positioning

博士 === 中原大學 === 機械工程研究所 === 105 === This study focuses on the development of appropriate control method for a 6DOF Stewart-type platform driven by piezoelectric actuators. It is aimed to preserve good positioning accuracy while encountered with external forces in particular. Kinematic calibration is...

Full description

Bibliographic Details
Main Authors: Nguyen-Van Tho, 阮文壽
Other Authors: Yung Ting
Format: Others
Language:en_US
Published: 2017
Online Access:http://ndltd.ncl.edu.tw/handle/64476929321522779193
Description
Summary:博士 === 中原大學 === 機械工程研究所 === 105 === This study focuses on the development of appropriate control method for a 6DOF Stewart-type platform driven by piezoelectric actuators. It is aimed to preserve good positioning accuracy while encountered with external forces in particular. Kinematic calibration is firstly carried out by using pose measurement to improve the accuracy of kinematic parameters. Negative factors of using piezoelectric actuators such as nonlinear hysteresis, creep, drifting disturbance, and temperature rise that directly affect the accuracy and steadiness of the system are concerned. In this article, modeling of the hysteresis of a piezoelectric actuator is derived to build a hysteresis feedforward controller by means of Preisach method that is able to deal with the rate-independent nonlinear hysteresis. Exponentially Weighted Moving Average (EWMA) method has been widely used in statistical process control and verified its capability of overcoming systematic change and drift disturbance. An attempt is to map the EWMA method into a run-to-run (RtR) Model Reference Adaptive System (MRAS) and combine with the hysteresis feedforward controller for position control. Similarly, a Predictor Corrector Control (PCC) with two stages of EWMA formulas is also used and verified its capability of overcoming the drifting disturbance due to creep and temperature dependence of piezoelectric actuators. Besides the position controller, an improved robust force feedback controller that is based on the idea of combining a nonlinear PID controller with the computed torque method is also investigated. The algorithm provides an essential way of dealing with unacknowledged interacting forces and variations of the environment characteristics. An example of surface grinding on different materials is investigated. Force sensor and capacitive displacement sensors are used to measure the interacting forces and the deformation respectively on the surface where the indentor tool is attached. Several case studies are performed to evaluate its effectiveness and robustness of the proposed controller. Comparison of the proposed controller with other controllers for free-space motion as well as for manipulation encountered with external load is carried out. The experiment results show that using the proposed composite controller is much better than using the feedforward or force feedback controller alone. The position accuracy can achieve root mean square error RMSE (x:95.436nm, y:172.513nm, z:1111.581nm) in translation and (θx:1.112nrad, θy:1.009 nrad, θz:0.689nrad) in orientation for free-space manipulation and RMSE (x:257.442nm, y:182.306nm, and z:1187.987nm) in translation and (θx:3.35nrad, θy:7.015 nrad, θz:0.687nrad) in orientation while encountered with external load.