Method for classifying electroencephalogram signals based on variable parameter recurrent neural network

A recurrent neural network and EEG technology, applied in the field of EEG signal classification based on variable-parameter recurrent neural network, can solve problems such as slow solution speed, achieve high robustness, improve signal-to-noise ratio, classification accuracy and high precision effect

Pending Publication Date: 2022-06-03
SOUTH CHINA UNIV OF TECH
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Although the support vector machine is easy to use, the traditional parameter solving method is a serial solution algorithm, and the solution speed is slow.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Method for classifying electroencephalogram signals based on variable parameter recurrent neural network
  • Method for classifying electroencephalogram signals based on variable parameter recurrent neural network
  • Method for classifying electroencephalogram signals based on variable parameter recurrent neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0089] figure 1 This embodiment is a flow chart of a method for classifying EEG signals based on variable-parameter recurrent neural network. The illustrated steps can complete the classification of EEG signals based on wavelet transform support vector recurrent neural network.

[0090] A method and system implementation of EEG signal classification based on variable-parameter recurrent neural network, comprising the following steps:

[0091] 1) Collect the user's EEG signal through the electrode cap, and then preprocess the data, including the determination of the time window, the selection of electrode channels, continuous wavelet transform and normalization;

[0092] 2) Based on the preprocessed data in step 1), using the principle of soft margin support vector machine, construct the convex quadratic programming problem of the motor imagery EEG signal classifier;

[0093] 3) Introduce the convex quadratic programming problem in step 2) into a penalty function to eliminate ...

Embodiment 2

[0163] Prompt the user to imagine a certain action (imagine the left foot and imagining the right foot) through the computer display screen, and then use the 3-channel electrode cap to collect the EEG signal 3-9s after the user accepts the prompt, and then preprocess the data, including determining The time window size of data sample collection is 3.25-6.25s, the electrode channel selection channel C3, channel C4, continuous wavelet transform and normalization processing. Among them, the continuous wavelet transform uses Morlet wavelet, which is expressed as:

[0164]

[0165] where exp( ) represents exponential function, t represents time, and then continuous wavelet transform is performed on the data, which can be expressed as:

[0166]

[0167] Where s is a scaling unit, s is selected from 7-30Hz, which is used to scale the wavelet, τ represents the center frequency, and x(t) is the input signal. Then use the max-min normalization method, which is expressed as:

[0...

Embodiment 3

[0227] Prompt the user to imagine a certain action (imagine the left foot and imagining the right foot) through the computer screen, and then use the Figure 4 The 22-channel electrode cap collects EEG signals 3-9s after the user accepts the prompt, and then preprocesses the data, including determining the time window size of data sample collection 3.25-6.25s, electrode channels, continuous wavelet transform and normalization processing . Among them, the continuous wavelet transform uses Morlet wavelet, which is expressed as:

[0228]

[0229] where exp( ) represents exponential function, t represents time, and then continuous wavelet transform is performed on the data, which can be expressed as:

[0230]

[0231] Where s is a scaling unit, s is selected from 7-30Hz, which is used to scale the wavelet, τ represents the center frequency, and x(t) is the input signal. Then use the max-min normalization method, which is expressed as:

[0232]

[0233] Where x represen...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention provides a method for classifying electroencephalogram signals based on a variable parameter recurrent neural network, which comprises the following steps of: collecting the electroencephalogram signals of a user through an electrode cap, and then preprocessing the data; constructing a convex quadratic programming problem of a motor imagery electroencephalogram signal classifier by utilizing a soft interval support vector machine principle; introducing a convex quadratic programming problem into a penalty function to eliminate inequality constraints; solving a convex quadratic programming problem with equality constraints through a variable parameter recurrent neural network; substituting the optimal solution of the convex quadratic programming problem into the classifier decision function to obtain the classifier decision function; and judging the category of the newly input motor imagery electroencephalogram signal through the obtained classifier decision function, and outputting the result. According to the method, the convex quadratic programming problem of the motor imagery electroencephalogram signal classifier is solved by using the variable parameter recurrent neural network, and the method has the advantages of high convergence speed, good robustness and high classification accuracy.

Description

technical field [0001] The invention belongs to the field of motor imagery electroencephalogram signal classification, in particular to a method for classifying electroencephalogram signals based on variable-parameter recursive neural network. Background technique [0002] Electroencephalogram (EEG) decoding of motor imagery is an important part of brain-computer interface (BCI), which can help patients with movement disorders to communicate with the outside world through external devices. Compared with the P300 EEG signal, the motor imagery EEG signal does not require the user to focus on strong stimulation information such as screen characters when collecting user signals, and the use burden is lower. It is conducive to the analysis and classification of EEG signals, so the classification of EEG signals for motor imagery has always been a hot field of research at home and abroad (S. Chaudhary, S. Taran, V. Bajaj, A. Sengur. Convolutional Neural Network Based Approach Towar...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/62G06K9/00G06N3/04G06N3/08
CPCG06N3/08G06N3/044G06F2218/02G06F2218/12G06F18/2411
Inventor 张智军陈广强任肖辉
Owner SOUTH CHINA UNIV OF TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap