DYNAMICAL MODELS OF REPEATED GOAL-DIRECTED MOVEMENTS

Joby John and Joseph P. Cusumano Nonlinear Dynamics Laboratory, Department of Engineering Science and Mechanics,
Pennsylvania State University, University Park, PA, USA,

INTRODUCTION A class of discrete dynamical systems for modeling repeated goal-directed kinematically redundant human movements is presented. The approach is based on the definition of movement tasks using goal functions and Goal Equivalent Manifolds (GEMs) (Cusumano and Cesari, 2006), which helps us to address “Bernstein’s degrees of freedom problem”. The example of a ball-throwing task is used to illustrate the central ideas. The resulting perception-action models have a hierarchical structure involving in-trial action templates and an inter-trial stochastic optimal error correction. The models identify the im-portant factors on which variability in repetitive skilled performance depends.