An Improved Calibration and Compensation Method for Lever-arm Errors between Two Rotational Inertial Navigation Systems

Abstract

To ensure the reliability of navigation system in marine and vehicle applications, the carrier is usually equipped with two or more rotational inertial navigation systems (RINS). The navigation information from different RINSs could be deeply integrated based on data fusion. During the data fusion process, the lever-arm errors between different RINSs would affect the velocity fusion accuracy. To improve the fusion accuracy, the lever-arm errors should be calibrated. The traditional calibration method is based on the acceleration errors model, which contains the angular velocity and the angular acceleration. However, the angular acceleration cannot be directly obtained from the outputs of RINS or other equipment. The general method for obtaining angular acceleration is to calculate the difference of the angular velocity, which would inevitably amplify the calculation errors and decrease the calibration accuracy. To solve this problem, an improved calibration method for lever-arm errors between two RINSs is proposed in this paper. Distinct with the traditional acceleration errors model and calibration method, the method could calibrate the lever-arm errors by using angular velocity of the carrier merely. The lever-arm errors model is deduced and improved. The calibration method is proposed based on the redundant configuration of two RINSs. The filter is established to achieve the optimal estimation of lever-arm errors. Experiments are carried out to prove and verify the proposed method and improved model. Results indicate that the improved calibration method could achieve the better calibration accuracy and the velocity differences of two RINSs are reduced obviously after the calibration results compensation, which illustrates the effectiveness and practical application values of the improved calibration and compensation method proposed in the paper.