Development and research of complex algorithm of inertial system for human motion parameters estimation

Abstract

Inertial motion capture is one most perspective technology for estimation of human motion parameters. Such systems use a network of inertial measurement units (IMUs) which are mounted on human body segments. To estimate kinematic motion parameters inertial motion capture systems (IMCS) is used data about body segments orientation. Algorithms which are used in IMCS are characterized by static and dynamic accuracy. Static accuracy of such systems is 0,2-0,5 degrees. But their dynamic accuracy during accelerating motions degrades to 2 degrees (RMS error). This study is intended for research and development of complex algorithm for work of one IMU of IMCS. Complex algorithm uses algorithm of strapdown inertial navigational system in geographical frame corrected by information of velocity and position data. These data are gotten using biomechanical skeleton model. The error level of such correction signals was estimated. A complex algorithm uses biomechanical velocities and positions to estimate velocity and position errors of data which are gotten from algorithm of strapdown system. These errors are used to form special correction signals in Poisson equation, equations for velocity and position calculation. It is shown effectiveness of proposed algorithm for kinematic human motion parameters estimation during accelerating segment motion. Maximal pitch and roll errors don’t exceed 0,8 degree. Positional RMS error is 0,04 m, velocity – 0,38 m/s. Such results show effectiveness of algorithm for estimation of segment orientation and position. Segment’s velocity signal has less error when it is gotten uses biomechanical skeleton model.