Optimal Sensor Fusion in Redundant Inertial Measurement Unit

In order to improve the reliability and accuracy of inertial navigation system in some navigation applications, the redundant inertial measurement unit (RIMU) is used. In this study, an optimal sensor fusion is introduced, and drift compensation based on this sensor fusion is presented subsequently. And the sensor fusion and drift compensation are combined into one algorithm. In this algorithm, faulty drift sensors are not isolated, but are compensated. So the redundant inertial navigation system can maintain system's redundancy. An errors model of RIMU was derived firstly, the optimal sensor fusion and the drift compensation algorithm were introduced secondly, and several simulations were carried out and proved effective of the proposed algorithm lastly.