BOUNDARY CONDITIONS IN HUMAN MOVEMENT II: THE FOURIER APPROACH

摘要

Time and frequency domains are merely alternative representations of the same underlying information, yet we argue that the frequency representation is more useful in understanding the control of fast human movements. We outline the deficiencies of temporal kinematic analyses and the advantages resulting from Fourier analysis. Using Monte Carlo simulations of noise-perturbed minimum jerk movements, we demonstrate the reliability of the spectral measurements up to a maximum frequency. The frequency limitation is determined by a combination of movement bandwidth, discontinuity order set by the boundary conditions, and spectral density of recording noise. For minimum jerk movements and current arm movement instrumentation this frequency is ～ 28Hz.