Fatigue state recognition method and system based on deep contraction sparse autoencoder network

Abstract

The present invention provides a fatigue state recognition method and system based on deep contraction and sparse self-encoding network, including: collecting the EEG signal of the person under test; using a filter to decompose the EEG signal to obtain the main components of the EEG signal in four different frequency bands , and then reorganized to obtain new EEG signals; build a deep contraction sparse autoencoder network, set the number of network layer nodes, extract features from the new EEG signals, and obtain abstract features; use the abstract features of the obtained EEG signals to perform Identify the fatigue state of the tested personnel, and obtain the fatigue state of the tested personnel. The invention can timely and accurately identify the fatigue state of the tested personnel, for example, remind the pilot before entering the fatigue state, which is conducive to improving flight safety and helping the flight organization to better arrange the pilot's load tasks.

Description

technical field [0001] The present invention relates to the field of EEG signal processing, in particular to a fatigue state recognition method and system based on a deep contraction sparse autoencoder network. Background technique [0002] As countries around the world pay more and more attention to aviation safety and advance in aircraft design and manufacturing technology, the reliability and safety of aircraft have been greatly improved, and the proportion of accidents caused by aircraft mechanical failure has dropped from 80% to 20%. , but the proportion of accidents caused by human factors is gradually increasing. Through investigation, it is found that the human factor that the National Security Agency of the United States has been most concerned about in the past forty years is pilot fatigue. According to its statistics, there are more than 300 aviation accidents caused by pilot fatigue. Pilots mainly show fatigue, drowsiness, lack of sleep and other states. Theref...

AI Technical Summary

Problems solved by technology

[0004] The current research proves that EEG signals can be used as the gold standard for human fatigue detection. Traditional EEG signal research mainly focuses on artificially extracting the characteristics of EEG signals, but this method is easily limited by the research level of experts themselves.

Images