Research on numerical compensation method of tandem force sensor installed at the end of industrial robot

Abstract

Tandem force sensor is an important sensor in human-robot collaboration and learning from demonstration, equipped with which robot can perceive the environment state and human attention simultaneously. The posture of a tandem force sensor changes with robot's posture, which leads to the output of the sensor cannot accurately reflect the environmental states and human intention. In order to compensate the measurement error caused by attitude change, the numerical compensation method of tandem force sensor based on deep learning is proposed in this paper. And the experimental results of numerical compensation show that this method can effectively eliminate the influence of attitude change on the measurement of the sensor.