Modeling the control strategies that humans use to control nonminimum-phase systems

This paper examines the control strategies that humans use to interact with an unknown nonminimum-phase system. We present results from an experiment in which 3 human subjects interact with an unknown nonminimum-phase system 40 times over a 1-week period. We use subsystem identification to model the control strategies that each subject uses during each trial. In particular, we identify feedback and feedforward controllers that model the subjects´ control strategies. The identified controllers suggest that the subjects learn to approximate and use the inverse plant dynamics (over a finite frequency range) in feedforward even though the plant is nonminimum phase.