A Method of Generating Virtual Fixture for Robot Machining

Abstract

The invention belongs to related technical field of intelligent robots, and discloses a generation method for robot processing virtual fixtures. The generation method comprises the following steps: the robot kinesthesis demonstration is performed on parts to be machined, and robot kinesthesis demonstration data sets are acquired; curved surface manifolds of the surfaces of the parts to be machinedare generated according to the acquired demonstration data sets; and virtual fixtures needed by tasks are generated according to the obtained curved surface manifolds. The invention further disclosesa corresponding process of machining complicated curved surfaces by a robot. Compared with the prior art, the generation method not only can generate needed complicated curved surface machining virtual fixtures with higher efficiency and higher precision, but also can fully use generalized characteristics of curved surface manifolds to build the virtual fixtures on similar parts, meanwhile, has the advantages of convenience to control, high stability and high adaptability, and prominently improves the machining quality of finally obtained robot workpieces.

Description

technical field [0001] The invention belongs to the technical field related to intelligent robots, and more specifically relates to a method for generating virtual fixtures for robot processing, which is especially suitable for applications where robots are used to process workpieces such as complex curved surfaces. Background technique [0002] With the continuous improvement of the industrial level, complex curved surfaces are more and more widely used in aerospace engine blades, wind power blades, and high-speed rail body-in-white. Due to its low cost, good flexibility, and high efficiency, robots are widely used in the above-mentioned complex surface grinding and polishing. At present, for difficult-to-program parts of complex surfaces such as damping tables, manual teaching is mainly used. The so-called robot teaching is the process of simulating manual actions through artificial teaching robots, and then specifying the actions that the robot should perform, that is, u...

AI Technical Summary

Problems solved by technology

[0004] Further retrieval analysis shows that, on the one hand, the generation of virtual fixtures in the prior art often relies on the existing geometric models, and the construction of virtual fixtures requires strong expert experience and a long design cycle; on the other hand, the generation of The virtual fixtures are often only applicable to the current parts, and cannot be applied to similar parts, and the generalization ability is poor

Images