Feature extraction and stability assessment of pedaling motion using three-axis accelerometer

Abstract

In the field of cycling competition, maximizing the performance of a cyclist is necessary to satisfy the following conditions: 1) Improving the pedaling skills that convert his/her physical strength to the impulsive force of a bicycle and 2) Setting the mechanical components of a bicycle to meet his/her physical properties. The relation between pedaling skill and mechanical components of a bicycle has not been discussed thoroughly for a long time. This study aims to propose a method to extract the pedaling motion with a three-axis accelerometer among different levels of cyclists and to assess stability from the mechanical information of the bicycle and extracted features by Kernel principal component analysis. In this paper, we concluded that a skilled cyclist switches their pedaling motion at a certain crank rotation angle; however, this is not the case with beginner and intermediate cyclists. These results suggested that cooperative motions of lower extremities contribute to a stable pedaling motion and optimize the efficiency of cyclists' pedaling motion.