The measurement step of the existing calibration approaches for robotic manipulators can take a considerable amount of time to settle a robotic manipulator down at certain static configurations, making the calibration approaches time-consuming. For applications of robotic manipulators requiring periodic recalibration (e.g., human–robot collaborative production lines and robotic inspecting systems), the time consumption of the data collection phase is a critical issue. This paper proposes a fast kinematic calibration approach for robotic manipulators, based on the measurement of a robotic manipulator tracking only a smooth and continuous time-optimal trajectory, rather than static measurement. Data samples on configurations are recorded continuously without settling the robotic manipulator down. To demonstrate and evaluate the proposed approach, experiments are performed based on a four degrees-of-freedom parallel manipulator. Experiment results show that compared to an existing calibration approach based on static measurement, the proposed approach improves the time efficiency of calibration by 93.13% with only a position accuracy loss of 1.77% and an orientation accuracy loss of 2.36%.
Journal of Mechanisms and Robotics Open Issues
