Geometric parameter identification of a dual-arm robot by using closed-chain constraint and optimization technique

Abstract

The positioning accuracy of the empirical robot manipulators is determined by various factors, such as kinematic accuracy, structure rigidity, and controller performance. Here, we report on the development of a new and straightforward technique to calibrate the kinematic parameters of a dual-arm robot under uncertainty. In comparison with other techniques, which generally rely on using other instruments to calibrate the manipulators, the proposed method utilizes the intrinsic characteristics of the dual-arm robot for calibration. In particular, when the two arms grasp each other, a formed closed chain can be operated as the constraint equation for the kinematic parameter optimization of the two arms. In the optimization process, the dual-arm robot has to pose in various configurations to yield better performance, and thus a motion generation strategy of the dual-arm robot is proposed, where one arm serves as the master to track the designated trajectory and the other arm serves as the slave to track the motion of the master arm by using a compliance control strategy. The proposed calibration method was experimentally validated, and the results confirm that the positioning accuracy of both arms can be improved.

Kinematic parameters calibration of industrial robot based on RWS-PSO algorithm

“…As the standard D-H method states that each link has four kinematic parameters, the errors between the theoretical values and actual values are defined as Dq = (Da i , Da i , Dd i , Du i )(i = 1, 2 Á Á Á 6), which are the parameter errors that need to be calibrated. When the kinematic parameter errors exist, the actual posture of the end point can be expressed as equation (3).…”

“…These changes may result in the emergence of kinematic parameter errors, which directly affect the modeling and calculation of industrial robots, and ultimately decrease the absolute positioning accuracy. 3 Currently, the absolute positioning accuracy is mostly greater than 1 mm, which is considered unacceptable in precision manufacturing. Therefore, it is essential to determine the kinematic parameter errors through calibration and compensate them to effectively decrease the posture error and increase the absolute positioning accuracy effectively.…”

A dual-arm manipulation strategy using position/force errors and Kalman filter

“…In contrast, the custom-made robot has thorough transparency in mechatronic and software architecture, so the control strategy was much easier to implement on the robot, and its performance was easier to evaluate without being contaminated by other factors that could not be controlled. The kinematic configuration of the dual-arm robot can be calibrated using its closed-chain configuration without using external sensors (Li et al, 2016; Yang et al, 2018).…”