Multi-space fusion man-machine skill migration and parameter compensation method and system

Abstract

The invention discloses a multi-space fusion man-machine skill migration and parameter compensation method and system, and the method comprises the steps: obtaining a multi-space skill parameter through human demonstration, decomposing a stiffness multi-space matrix into a feature value matrix and a feature vector matrix, converting a Riemannian space matrix into a quaternion, and representing the quaternion in an axial angle form; a multi-space fusion dynamic system, a regular system module and a linear weighted regression updating module are used for achieving skill parameter migration, the skill parameters are migrated to a robot humanoid variable impedance controller, and the robot reproduction effect obtained by comparing the corresponding skill parameters with the effect demonstrated by a demonstrator under the skill parameters. And importing the difference value under multiple demonstrations and the actual operation effect of the robot into a probability model for training, establishing a rule model between the actual value and the difference value, and performing skill parameter compensation according to the difference between predicted values obtained by robot end feedback and rule model regression in an actual scene. According to the invention, multi-space man-machine skill transmission can be realized, and an accurate reproduction effect is ensured.

Description

technical field [0001] The invention belongs to the technical field of fusion robots, and in particular relates to a multi-space fusion human-machine skill transfer and parameter compensation method and system. Background technique [0002] Traditional robots, especially industrial robots, mostly achieve relatively single tasks through programming and teaching by professional technicians. This type of robot plays an important role in a structured environment. It can replace humans to complete simple and repetitive tasks, greatly reducing the work intensity of humans. It has the advantages of high positioning accuracy, high work efficiency, and large output force. It has been widely used in medium and large enterprises with fixed production methods such as logistics, automobile industry, aircraft assembly, port and wharf, and has brought considerable work efficiency and economic benefits. However, such robots still have shortcomings such as poor universality, low flexibility...

AI Technical Summary

Problems solved by technology

At present, the existing research on human-computer skill transfer technology mainly focuses on the Euclidean space, and rarely takes the Riemannian space into consideration.

Moreover, skill transfer is mainly aimed at kinematic parameters, and few studies have given dynamic skill parameters such as the stiffness characteristics of the demonstrator and the force at the end of the operation process to the robot.

Images