EEG-based quality of teleoperator identification using emotional states model

Fatigue recognition has been a point of interest in many fields including robotic teleoperation. In our previous work we developed a method for evaluating the fatigue state of teleoperator using Quality of teleoperator (QoT). QoT can be estimated based on Radial Basis Function (RBF) network with 5 QoT indicators as input variables namely, long term excitement, short term excitement, boredom, meditation, and frustration. In this paper a new approach is presented for measuring the QoT based on EEG signals which is composed of two stages. In the first stage the EEG signals is projected to the Pleasure, Arousal, and Dominance (PAD) emotional states model. The resulted three emotional states then projected to QoT single dimension scale in the second stage. The experimental results showed improvement in the cross validation error, reduced form 10.7891 to 9.6133, and the complexity of the model is reduced by minimizing the number of variables from five to three. The statistical analysis results show positive correlation between pleasure and dominance states with the QoT value (p<;0.05) and insignificant correlation between arousal state and QoT value.