Extended Kalman Filter for Magnetometer-Free Estimation of Spacesuit Wearer Joint Kinematics Using Inertial Measurements

Applications of inertial sensing to the study of human motion in ambulatory environments often suffer from a dependence on magnetometers, which are susceptible to nonuniform and time-varying magnetic disturbances. The current work presents an inertial sensor fusion algorithm that will enable magnetometer-free inertial sensing of human joint kinematics in ambulatory environments. The particular motivating application for the current work is the need to study human motion inside a spacesuit in order to inform spacesuit design. The algorithm synthesizes several techniques for optimally fusing inertial sensor measurements in order to reduce drift rather than correcting it. A peak attitude estimation error of 2 degrees over 35 minutes is demonstrated in simulation.