DEVELOPMENT OF THE 6-AXIS FORCE/MOMENT SENSOR

• Main Author: Anton Royanto Ahmad
• Other Authors: Chyi-Yeu Lin
• Format: Others
• Language: en_US
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/51087760130602474572

碩士 === 國立臺灣科技大學 === 機械工程系 === 103 === The six-axis force sensors have been widely used and become very important components in robot-based intelligent automation. Structure design, strain gauge arrangement, and calibration are successfully integrated to obtain an accurate and decoupled six-axis force sensor design. In order to develop economical decoupled six-axis force sensor, Maltese cross type is chosen in this study. A new arrangement of strain gauge is applied and double parallel linear strain gauge is used. Two calibration methods for this novel arrangement are artificial neural network and polynomial equation. Experimental results on single force and single moment show the artificial neural network approach performs better than the polynomial equation approach. However, experimental results on combined forces and moments situations show that the polynomial equation approach performs better than the artificial neural network approach.