High-Precision and Efficient Calibration of Robot Polishing Systems Using an Adaptive Residual EKF Optimized by MIPO

Area of Science:

• Robotics
• Control Systems
• Optimization Algorithms

Background:

• Standard Extended Kalman Filters (EKF) suffer from accumulated truncation errors and sensitivity to noise parameters.
• Accurate kinematic calibration is crucial for high-precision robotic polishing systems.
• Existing methods often lack robustness and efficiency in real-world applications.

Purpose of the Study:

• To develop an improved Kalman filter method for accurate and efficient kinematic parameter calibration in robotic polishing.
• To address limitations of traditional EKF, including error accumulation and noise parameter sensitivity.
• To enhance the robustness and reduce computation time of robotic calibration processes.

Main Methods:

• Proposed an adaptive residual extended Kalman filter (ARKEKF) integrated with a multi-strategy improved parrot optimization algorithm (MIPO).
• Introduced a gradient stabilizer to mitigate estimation degradation from truncation errors.
• Employed MIPO for adaptive optimization of process and measurement noise covariance matrices.

Main Results:

• The MIPO-ARKEKF method reduced root mean square positioning error by 45.58% on a KUKA robot.
• Achieved comparable accuracy to existing methods but with 34.88%–65.08% less computation time.
• Demonstrated improved end-effector trajectory tracking and polishing quality in experiments.

Conclusions:

• The MIPO-ARKEKF method offers a significant improvement in kinematic calibration accuracy and efficiency for robotic polishing.
• The adaptive optimization and error mitigation strategies enhance robustness under practical measurement uncertainties.
• This approach provides an efficient solution for high-precision robotic polishing tasks, improving overall quality.