Neural-Network-Based Six-axis Force/Torque Robot Sensor Calibration

Abstract

Six-axis force/torque robot sensor is an important component of intelligent robots. It is unable to interpret the relationship between input and output accurately by means of the conventional least-squares method for six-axis force/torque sensor calibration, because the sensor may suffer from non-linearity and various forms of uncertainty. In this paper, neural-networks method is used for the robot sensor calibration. This method and the least-squares method are both presented in this paper. And the results of both methods are compared and discussed. The results show that neural-networks-based calibration is more efficient and accurate.