Predictor-based Control of Excitement as Human Response Signal to a Dynamic Virtual 3D Face

This paper introduces the application of predictor-based control with constraints of human response to a dynamic virtual 3D face. We are using changing distance-between-eyes in a woman 3D face as a stimulus - control signal. Human response to the stimulus is observed using EEG-based excitement signal - output signal. The technique of dynamic systems identification which ensures stability and possible higher gain of the model for building a predictive input-output model of control plant is applied. Two predictor-based control schemes with a minimum variance or a generalized minimum variance control quality and constrained control signal magnitude and change rate are developed. High prediction accuracy and control quality are demonstrated by modelling results.