Estimation of orientation and position of the manipulator using the extended kalman filter based on measurements from the accelerometers

Authors present the kinematic structure of measurement arm along with its construction for efficient estimation of orientation and position of the manipulator using extended Kalman filter. The major innovation of the arm is that it only uses accelerometers as gravity sensors for determining relative positions of the links. This article presents the problem of position estimation based on measurements with high noise and the use of the extended Kalman filter to limit the impact of noise on the measurement. Repeatability tests were performed using custom made test stand.