MENTAL WORKLOAD CLASSIFICATION BASED ON VISUAL AND OPERATIONAL EEG SIGNALS

The degree of mental workload directly affects the accuracy and safety of the task in the human-computer operating system, so it is very meaningful to study the state of mental workload of the operator. The common classification methods of mental workload direct uses EEG features to classify, which has low accuracy. This paper proposes a classification method with high accuracy and reliability for the mental workload classification of visual and operational task. This method directly extracts the energy characteristics of four different frequency bands from the independent components of the EEG, and then classifies them. The research results show that the accuracy of the proposed method is improved by 25.62%.