Assessment of GPS/MEMS-IMU Integration Performance in Ski Racing

This paper describes an experiment involving a professional downhill skier equipped with GPS and MEMS-IMU, as well as a triad of magnetometers and a tactical-grade inertial unit. The experiment illustrates the navigation performance of GPS/MEMS integration compared to high-quality GPS/INS integration challenged by the environmental conditions and dynamics of ski-racing. The MEMS sensors’ output is directly compared to the reference signals. This information is used to validate the error models of the inertial and magnetic measurements and to select the appropriate integration algorithms. INTRODUCTION Today, material development and testing in Alpine skiing is based on repeated measurements with timing cells or in wind tunnels; the performance analysis of athletes on simple chronometry and video recordings. These methods appear to be vulnerable to meteorological conditions and the fatigue of the test persons. Furthermore, current methods based on videogrammetry [1] and approaches based on timing cells don’t offer a flexible choice of the intermediate times and the analysis of accelerations and velocities along the whole track. Therefore, new methods offering precise determination of the skiers’ position, velocity and attitude are wanted. Satellite-based positioning has already proven its effectiveness in car racing [2], rowing [3] and Alpine Skiing [4]. The latter used dual-frequency GPS receivers to reconstruct the skiers’ trajectories. Because of ergonomic and economic restrictions, theses receivers will be reserved to few Sports applications with higher accuracy needs. [5] has shown that L1 only receivers were capable to determine the peak velocity of the speed skier with similar accuracies than dualfrequency receivers. [6] has demonstrated that the timing accuracies provided by L1 GPS chronometry achieved the same level of accuracy as L1/L2 GPS chronometry and traditional chronometry based on timing cells provided that the carrier-phase ambiguities were solved. However, considering the high dynamics of a skier and the ergonomic requirements placed on the equipment, today’s technological limits in GPS positioning are quickly reached or even exceeded [7]. Furthermore, the athlete’s environment is quickly alternating between open spaces and areas that are adverse to the reception of satellite signals (sudden satellite masking) which makes the carrier-phase ambiguity resolution difficult or even impossible. Furthermore, athletes and coaches are not only interested in the trajectories and velocities, but also in the motion analysis of segments of the human or the attitude of his equipment [8]. Therefore, this paper introduces a system of low cost Micro-Electro-Mechanical System (MEMS) Inertial