Intelligent skill training and evaluation system and method based on artificial intelligence

Abstract

The invention discloses an intelligent skill training and evaluation system and method based on artificial intelligence, and particularly relates to the field of intelligent skill training, and the method comprises the steps: inputting data collected by a sensor network, constructing a skill behavior channel set in a unified manner, and extracting a key behavior channel according to the influence intensity of each channel on the motion driving degree in task execution; generating a time sequence behavior tensor set under a time sequence dimension based on the key behavior channel, taking the time sequence behavior tensor set as an input expression of a skill continuous behavior, introducing an induction assembly including speed disturbance, attitude disturbance and path deviation disturbance into the time sequence behavior tensor set, and constructing a disturbance excitation space; the disturbance excitation space and the behavior tensor set are combined step by step. By actively constructing a disturbance excitation space and introducing various disturbance conditions, a misjudgment mechanism of data convergence, namely skill mastering, in traditional evaluation is broken, false stability caused by path dependence can be identified, and it is ensured that the real coping ability of a trainer is captured truthfully.

Images