Fast calibration method for industrial robot kinematics based on motion capture system

Area of Science:

• Robotics
• Mechatronics
• Control Systems

Background:

• Industrial robots (IRs) require precise kinematic calibration for optimal performance.
• Existing calibration methods are often time-consuming and unsuitable for dynamic environments.
• Accurate kinematic models are crucial for tasks like obstacle avoidance.

Purpose of the Study:

• To develop a fast and accurate kinematics calibration method for industrial robots (IRs).
• To address the limitations of existing methods in complex and temporary IR-related scenarios.
• To improve obstacle avoidance capabilities by considering intermediate link positions.

Main Methods:

• Dividing IR kinematic calibration into four specific application situations.
• A two-step calibration process: determining joint axes for rough parameters, then refining with stochastic gradient descent.
• Introducing obstacle-avoidance auxiliary points to account for intermediate link positions.

Main Results:

• The proposed method achieves fast and accurate calibration in complex scenarios.
• Demonstrated effectiveness through simulations and experiments using a motion capture system.
• Successfully improved obstacle avoidance considerations compared to traditional methods.

Conclusions:

• The developed method offers a significant advancement in rapid industrial robot calibration.
• The approach is particularly effective for complex and temporary operational environments.
• The inclusion of obstacle-avoidance points enhances the practical utility of the calibration.