Psycho-Physiologically-Based Real Time Adaptive General Type 2 Fuzzy Modelling and Self-Organising Control of Operator's Performance Undertaking a Cognitive Task

摘要

—This paper presents a new modelling and control fuzzy-based framework validated withudreal-time experiments on human participants experiencing stress via mental arithmetic cognitiveudtasks identified through psycho-physiological markers. The ultimate aim of the modelling/controludframework is to prevent performance breakdown in human-computer interactive systems with audspecial focus on human performance. Two designed modelling/control experiments which consist ofudcarrying-out arithmetic operations of varying difficulty levels were performed by 10 participantsud(operators) in the study. With this new technique, modelling is achieved through a new adaptive,udself-organizing and interpretable modelling framework based on General Type-2 Fuzzy sets. Thisudframework is able to learn in real-time through the implementation of a re-structuredudperformance-learning algorithm that identifies important features in the data without the need forudprior training. The information learnt by the model is later exploited via an Energy Model BasedudController that infers adequate control actions by changing the difficulty level of the arithmeticudoperations in the human-computer-interaction system; these actions being based on the most currentudpsycho-physiological state of the subject under study. The real-time implementation of the proposedudmodelling and control configurations for the human-machine-interaction under study shows superiorudperformance as compared to other forms of modelling and control, with minimal intervention in terms of model re-training or parameter re-tuning to deal with uncertainties, disturbances andudinter/intra-subject parameter variability.