Position-Based Robot Calibration and Compensation Using an Improved Adjoint Error Model

The original adjoint error model is a complete, continuous, and minimal calibration model. However, the original model is defective in its mathematical derivation, and the method of updating kinematic parameters is also complicated and indirect; meanwhile, the orientation measurement of the end-effector is very difficult to implement. To address these problems, we present a complete and unified model for robot calibration based on an improved adjoint error model, in which only the position measurements of the end-effector are required. In addition, a compensation algorithm is proposed to improve the robot position accuracy using the calibration results; the proposed algorithm does not require modification of the kinematic parameters of the robot controller and can be applied to robots with different degrees of freedom. Simulations on a PUMA560 robot and a SCARA robot were performed to validate our algorithms. Furthermore, the proposed algorithms were applied to our self-designed modular robots with four and six degrees of freedom. The experimental results show that the average accuracy of the robots is enhanced by approximately one order of magnitude after compensation.