Self-adaptation evaluation device and method for mental fatigue

Abstract

The invention discloses a self-adaptation evaluation method for mental fatigue. The method includes the steps that 1, electroencephalogram signals are collected to form electroencephalogram samples; 2, low-pass filtering is performed on the collected electroencephalogram signals, four different electroencephalogram rhythm signals of delta, theta, alpha and beta are extracted, and the relative energy and the energy ratio of the three rhythm signals of delta, theta and alpha serve as feature information of all channel electroencephalogram signals; 3, trained initial basis classifiers are utilized for predicting the electroencephalogram samples collected on line, predicting results are selected through a strategy of majority voting, the samples are marked, and the evaluation results for a current fatigue state are given. By the utilization of a selective integration technology, a set of basis classifiers with high speed and difference is trained to serve as the initial classifiers for a semi-supervised evaluation mode, and is updated in parallel in an on-line mode to be integrated into a strong classifier, in this way, it is guaranteed that the generalization capacity of the classifiers is enhanced, and meanwhile high classification precision and high operation speed are achieved.

Description

technical field [0001] The invention belongs to the interdisciplinary field of biomedicine and informatics, and in particular relates to an online mental fatigue assessment device and method based on integrated learning and semi-supervised learning. Background technique [0002] Due to the high pressure of social competition, mental fatigue has gradually become one of the main reasons affecting human health. Mental fatigue can cause changes in people's physical and psychological states, leading to a decline in alertness, sustained attention, working memory, judgment, and decision-making. In modern society, especially in real-time monitoring, transportation, high-risk operations, aerospace and other operations, if the staff is in a state of fatigue, it is very easy to perform random operations and violations, which will lead to safety accidents. Therefore, it is very necessary to study objective, reliable and accurate mental fatigue detection methods for preventing safety ac...

AI Technical Summary

Problems solved by technology

And when the number of labeled training samples is small, the generalization ability of the obtained classifier is often poor

Furthermore, classifiers built during training are fixed during work and are not adaptive

Images