Design and optimization of a joint torque sensor with small crosstalk error for robots. (15th August 2022)

• Record Type: Journal Article
• Authors: Fu, Bing; Cai, Ganwei
• Is Part Of: Measurement. Volume 200(2022)
• Publication Date: 2022-08-15
• Subjects: Joint torque sensor -- Structural optimization design -- Crosstalk -- Collaborative robot; Weights and measures -- Periodicals; Measurement -- Periodicals; Measurement; Weights and measures; Periodicals; 530.8
• DOI: 10.1016/j.measurement.2022.111629
• Languages: English
• ISSNs: 0263-2241

Abstract

Highlights: A novel joint torque sensor was designed with good anti-crosstalk performance. A multi-objective optimization design method for joint torque sensor was proposed. An anti-crosstalk circuit of the strain gauge-based joint torque sensor was proposed. Smaller crosstalk error makes expensive cross roller bearings unnecessary for joint torque sensor. An economical and practical design and calibration method for the joint torque sensor. Abstract: Joint torque sensors (JTS) are often mounted on the joints of a collaborative robot to measure torque but encounters crosstalk problem that affect measurement accuracy. We propose a multi-objective optimized design that considers sensitivity, crosstalk, and stiffness, and develop an anti-crosstalk circuit of a strain gauge-based unit make up a novel JTS. Experimental results show that the maximum crosstalk error was 0.429 Nm under a moment load of 54.384 Nm, 0.274 Nm under a radial load of 295.123 N, and 0.123 Nm under an axial load of 295.123 N without using cross-roller bearings. The proposed JTS can properly deal with crosstalk, and can accurately and directly measure the torque of interest without requiring expensive cross-roller bearings. This simplifies the structural design and reduces the weight and the cost of the robot. These results verify the efficiency and economy of the proposed method of customizing the JTS.