Parallel nesting and autonomous preferential classifier for intelligent artificial limb brain-myoelectricity fusion perception

Abstract

The invention discloses a parallel nesting and autonomous preferential classifier for intelligent artificial limb brain-myoelectricity fusion perception. A multi-convolution neural network classifieris constructed, time span characteristic changes of electroencephalogram, myoelectricity and brain-myoelectricity fusion characteristics are considered, brain-myoelectricity weight indexes are constructed by measuring brain and myoelectricity activity degrees and characteristic levels, and the brain-myoelectricity weight indexes participate in construction and training of the classifier, so that the classifier can autonomously and intelligently adapt to a time span brain-myoelectricity fusion signal perception recognition task, and autonomous decision of optimal classification is realized. Compared with a traditional classifier which is simple in signal source and does not have an intelligent preferential function, the classifier has more excellent signal analysis performance and self-adaptive capability, and is suitable for equipment with time span model updating requirements such as rehabilitation artificial limbs.

Description

technical field [0001] The invention relates to the field of brain-computer interface technology and artificial intelligence, in particular to a parallel nesting and autonomous optimal classifier for brain-myoelectric fusion perception. Background technique [0002] In the electromechanical system, EEG signals and EMG signals contain physiological signals about action intentions and physical behavior characteristics, and are widely used in the fields of handicapped (rehabilitation) prostheses, exoskeletons, etc.; for disabled patients, brain signals representing action intentions Electrical signal recognition has always faced the problems of low accuracy, poor stability, and weak robustness, especially for unilateral limb movement manipulation, where the EEG intention to distinguish different movements comes from the motor sensory area of ​​​​the unilateral brain, and the degree of confusion is high; Myoelectric signals have more obvious characteristics than EEG signals, but...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to provide a parallel nesting and autonomous optimal classifier for intelligent prosthetic brain-myoelectric fusion perception, which solves the problems of the existing brain-myoelectric fusion perception strategy and rehabilitation intelligence

Images