Indoor Pedestrian Positioning Method Based on Pelvic Motion Model Using Inertial Sensors

Indoor positioning technology has been widespreadly studied by researchers at home and abroad. The sensor located at pelvis is used to estimate the location of human by acceleration. The existence of the acceleration error will lead to the increasing position error after double integrations. Step length algorithm using foot-mounted inertial sensor need to use Zero Velocity Update (ZUPT) method to reset velocity in order to reduce accumulated error. In this paper, we propose a new position method based on the pelvic motion model. The center of the pelvic is selected as the root point. The five sensors are located at the shank, thigh and the pelvis to measure the kinematics of the lower limbs. First, the coordinate values of the left and right knee joints are calculated by the rotation matrix. The roll obtained from the sensor attached to the shank is used to estimate the horizontal distance from the ankle to the knee joint. The yaw is used to estimate the heading of pedestrian. Finally, the position is determined by step length and heading information. The proposed method and the conventional method are tested in two scenarios: rectangular motion and straight-line motion. The experimental results show that the proposed position estimation method has higher position accuracy indoors.