A Noncontact Control Strategy for Circular Peg-in-Hole Assembly Guided by the 6-DOF Robot Based on Hybrid Vision

Abstract

Facing some special operating environments or conditions, existing control methods for the peg-in-hole assembly guided by robots always have their own disadvantages, for example, low efficiency or poor adaptability. For the above problem, in this article, a new circular peg-in-hole assembly control strategy is proposed for the 6-Degree of Freedom (DOF) robot based on hybrid visual measurements, avoiding peg-in-hole contacts during the robotic operation. In the strategy, the pose of the monocular camera mounted at the end-effector is adaptively adjusted to improve the image quality through an algorithm based on the rough pose measurement of the target hole by the binocular camera; the accurate 3-D pose of the hole is determined by an algorithm based on processing of high-quality images and the compensation of the orientation error. Combined with the robotic collision-free path planning, the automatic peg-in-hole assembly can be implemented in real setting. The assembly precision of the robotic system based on the proposed method is validated and discussed based on experimental results. Then, the minimal peg-in-hole interval relative to the alignment error is modeled through the spatial relation analysis to analyze the applicable condition of the robotic system with the control strategy. Also, the reliability of the proposed strategy is verified through experimental tests under some applicable conditions. Finally, suggestions and plans of future works are discussed for further extension of the application area of the proposed strategy, such as fields of precision and ultraprecision manufacturing. This contribution has the major significance on the automatic peg-in-hole assembly under 3-D operating environment.