LMD entropy feature and LVQ neural network-based motor imagination electroencephalogram signal identification method

Abstract

The invention discloses an LMD entropy feature and LVQ neural network-based motor imagination electroencephalogram signal identification method, and belongs to the technical field of brain-computer interfaces. The method comprises the steps of decomposing an electroencephalogram signal by applying an improved LMD method to obtain a series of PF components and residual errors of the original signal, and screening out the PF component containing a main characteristic frequency; calculating an energy entropy, a fuzzy entropy and a multi-scale entropy for the selected effective PF component, and fusing the three entropies into eigenvectors; and finally performing classification identification on the eigenvectors by using an LVQ neural network. According to the improved LMD method in the identification method, a self-extension method is adopted for improving a possibly existent end effect problem. The identification method solves the LMD end effect problem, can effectively perform electroencephalogram signal feature extraction, realizes effective identification of a motor imagination mode, and has a certain practical value.

Description

technical field [0001] The invention belongs to the field of electroencephalogram signal identification and analysis, and in particular relates to a motor imagery electroencephalogram signal identification method based on improved LMD, energy entropy, fuzzy entropy, multi-scale entropy and LVQ neural network. Background technique [0002] Brain Computer Interface (BCI) is a way for the human brain to communicate with the outside world, and it can be realized without relying on brain nerves and muscles. Brain-computer interface has bright application prospects in medical rehabilitation, games, military and other fields, and is a cutting-edge science in the 21st century. There are implantable and non-implantable methods for collecting brain bioelectrical signals, among which the non-implantable brain-computer interface based on EEG is widely used. Electroencephalogram (Electroencephalogram, EEG) reflects the functional state of the brain and the electrical activity of brain t...

AI Technical Summary

Problems solved by technology

However, for EEG signals with complex mechanisms, it is usually impossible to obtain accurate prior information, and it lacks the ability to adapt to different individuals

Empirical Mode Decomposition (EMD) is an adaptive signal analysis method, which decomposes EEG signals into multiple IMF components, but EMD decomposition will cause phenomena such as endpoint effects and modal aliasing, and the decomposition of EMD time consuming

Images