Low-Cost Mirror-Based Self-Calibration With Kinematic Error Compensation for Vision-Guided Composite Mobile Robots

Accurate calibration is crucial for improving the performance of composite mobile robots in large-scale assembly tasks. In this article, we propose a low-cost mirror-based self-calibration method designed to achieve high-accuracy calibration without the need for external measurement equipment. The proposed method simultaneously enables tool center point (TCP) calibration, hand-eye calibration, and mobile platform-arm calibration using reflective measurements from a planar mirror. A novel measurement strategy is introduced, which overcomes the limitations of traditional contact-based methods and addresses scenarios lacking well-defined reference points. Furthermore, an error compensation strategy based on differential kinematics is proposed to mitigate the effects of robot kinematic error and positioning errors. Unlike conventional calibration methods that rely on laser trackers or coordinate measuring machines, the proposed approach eliminates the need for additional instrumentation, offering a new, cost-effective measurement solution. Simulation and experimental results demonstrate an assembly accuracy of 0.212 mm, which meets the aerospace requirement of ≤0.300 mm for high-lock nut assemblies, ensuring reliability in aircraft panel manufacturing.