General measurement model of distributed accelerometers for inertial navigation

Abstract In some inertial navigation systems (INS), not all accelerometers are placed at the center of the body frame; they are called distributed accelerometers. Such accelerometers can sense both translational and rotational motions. A general measurement model is proposed to extract and combine all the translational and rotational information optimally. It is a state space model, with its 12 states being: positions, velocities, attitudes and angular rates. States can be calculated through Kalman filter. The proposed method is independent of sensors’ stochastic models and empirical kinematics models, quite the same as traditional INS. With the proposed model, other relevant data can be combined optimally.