An Improved Robust Method for Pose Estimation of Cylindrical Parts with Interference Features.

Area of Science:

• Aerospace Engineering
• Robotics
• Metrology

Background:

• Horizontal docking assembly is critical in aerospace manufacturing, requiring high precision and automation.
• Current measurement techniques face challenges with surface interferences, impacting accuracy.
• Intelligent measurement and adjustable support systems are essential for advanced aerospace assembly.

Purpose of the Study:

• To develop an intelligent measurement method for precise pose estimation in horizontal docking assembly.
• To enhance the robustness of pose measurement against surface interferences.
• To improve automation and precision in aerospace assembly processes.

Main Methods:

• Utilizing a laser scanning approach to acquire 3D point cloud data.
• Employing section profile fitting, including ellipse fitting and spatial straight line fitting, for pose parameter determination.
• Implementing a robust method combining M-estimation and RANSAC to mitigate interference effects and validate planar profiles.

Main Results:

• Accurate determination of part pose parameters from laser scanning data.
• Successful enhancement of measurement robustness against surface interferences.
• Validation of the proposed method through prototype fabrication and experimental testing.

Conclusions:

• The proposed laser scanning and section profile fitting method provides a robust solution for precise pose estimation in aerospace horizontal docking assembly.
• The developed robust technique effectively handles surface interferences, improving measurement reliability.
• This advancement contributes to higher automation and precision in aerospace manufacturing.