COMPARISON OF BIOMECHANICAL MOVEMENT PATTERNS BY MEANS OF ORTHOGONAL REFERENCE FUNCTIONS

Abstract

Common biomechanical analyses compare movements with the help of variables' intensities on certain instances. The analysis of variables' time courses during a single movement are mostly done just qualitatively or by singular variables. A quantitative comparison of movement patterns in terms of multivariate time courses could rarely be found in sports sciences. The present investigation demonstrates a method to compare sets of time courses of biomechanical variables. Eight throwing movements of a discus thrower during a learning process were filmed with two perpendicular positioned 16mm-cameras (LOCAM). From the digitized coordinates 20 body related variables were derived. The athletes' biomechanical movement pattern was defined by the description of the variables' time courses. The variables describe the joints' movement with angles and angular velocity. The time courses of the eight throws during one year are compared with each other. First the time courses and a set of six reference functions are normed to equal length and equal number of measurement points. The reference functions are six orthogonal mathematical functions. Each variables time-course is correlated with each of the six functions. The similarity of two variables' time courses is defined by a variables' similarity coefficient. With the help of the time courses of the variables' similarity coefficients, some variables can be determined which show synergies and dependence from the applied training contents. A few variables' similarity coefficient have an almost chaotic, hard explainable, time course during the learning period. The average of all variables' similarity coefficients yields to an overall similarity of the discus movement patterns. The time series of the overall similarity shows no resemblance to any variables' similarity time course. Obviously, the way how the whole biomechanical movement pattern changes during the learning process is different from the way that any single variables' time courses do. The changes of the movement patterns seem to have special dynamics.