Fatigue Detection Algorithm Based on Discrete Wavelet Transform of EEG Signals

EEG signals are usually used to study brain functions such as cognition, perception, emotion, and sleep, but the collected EEG signals are usually mixed with interference signals such as ocular artifacts. To solve this problem, this paper proposes an adaptive ocular artifacts removal algorithm that combines empirical mode decomposition (EMD), independent component analysis (ICA) and sample entropy (SampEn). In order to detect the fatigue state of the human body, this paper proposes a fatigue detection algorithm based on discrete wavelet transform (DWT). Extract the EEG rhythm wave, reconstruct the Theta wave, Alpha wave and Beta wave, and use the ratio of (Theta + Alpha)/Beta as the fatigue index to measure the fatigue degree of the human body. In 100 independent repeated tests, 91 times of the correct prediction of the state of the subjects, the recognition accuracy of the algorithm reached 91%, and the recognition accuracy of this algorithm reached 91%.