A measurement method for calibrating kinematic parameters of industrial robots with point constraint by a laser displacement sensor

Abstract

This paper proposes a novel calibration method for robot kinematic parameters by constraining a command point. At first, the profile errors derived from measuring the sphere surface by a laser displacement sensor are estimated and analyzed. A three-step measurement procedure is performed around a command point with only one single laser displacement sensor. Then, a kinematic error model including both the robot parameter errors and the errors between the robot base frame and the measurement frame is established. An identification algorithm comprising the least square method and the particle swarm optimization algorithm is developed. Comparison simulations are carried out with the proposed method, a least square method, and the traditional identified method. Experiments show that the estimated compensation results are significant, with the mean reduced from 10.259 mm to 0.845 mm. A verification experiment shows the mean of the relative deviations is decreased from 1.674 mm to 0.538 mm.