Sensorless Impedance Control of Delta Robot

碩士 === 國立臺灣科技大學 === 機械工程系 === 107 === With the increased utilization of industrial manipulators, the force control is gaining more and more attention and appreciation. The force control is also known as the compliant control. Among all proposed techniques, impedance control is the most commonly used...

Full description

Bibliographic Details
Main Authors: Han Guo, 郭翰
Other Authors: Meng-Kun Liu
Format: Others
Language:zh-TW
Published: 2019
Online Access:http://ndltd.ncl.edu.tw/handle/fga784
Description
Summary:碩士 === 國立臺灣科技大學 === 機械工程系 === 107 === With the increased utilization of industrial manipulators, the force control is gaining more and more attention and appreciation. The force control is also known as the compliant control. Among all proposed techniques, impedance control is the most commonly used method. The compliant control is very useful and effective especially in the situation of interaction with robots and humans, but it fundamentally requires installation of the force sensor to obtain the external force value as feedback signal. This drastically increases the cost of the implementation of such a haptic manipulator. Therefore, many researchers have proposed estimation methods to estimate the external force. Those estimation methods ease the requirement of the force sensor. The disturbance observer (DOB) is the most widely used, but has mainly applied to the series manipulator or low D.O.F. robot arms. On the other hand, the effectiveness and the performance of the force estimation applying to parallel manipulators are still not been investigated thoroughly. In this study, the external force estimation performance of DOB on the Delta robot is not as expected. Thus, a sensor-less impedance control method based on system identification (SID) and the concept of inverse objective impedance is proposed and applied. The proposed method was verified via simulation and experiments with satisfactory performance. This control scheme also applied to Delta robot for collision detection with pleasing result.