Error Modeling And Compensation Of Single-axis Rotation FOG SINS

As the rotating modulation technology has been successfully applied in optical fiber inertial navigation system, fiber rotary strap-down inertial navigation system gradually becomes a hot research topic in the field of navigation. In order to improve the performance of single-axis rotary fiber inertial navigation system, this paper focuses on the error of this system to do some further research. Through the identification and compensation of inertial device error, the characteristics analysis and calibration of the system error and the adoption of high precise initial alignment scheme, the navigation precision of the system is improved.First, the thorough analysis on the error of inertial device is carried on. As we know, the output error of fiber optic gyroscope, a core in fiber inertial navigation system, has a direct impact on the performance of navigation system. Due to the clear characteristics of the certainty error of fiber optic gyro and the relatively complex characteristics of the random one, the latter one is thoroughly analyzed in this paper. Based on the analysis on the evaluation method of the random error of fiber optic gyro, it is concluded that the least squares random error fitting in Allan variance analysis method has defects of distortion. Aiming at this problem, a kind of Allan variance fitting method based on model identification is put forward. By doing qualitative analysis before quantitative calculation, the precision of the random error analysis on fiber optic gyro is improved, which provides the basis of the compensation of the random error for inertial components.In order to further improve the accuracy of fiber inertial navigation system, the characteristics of system error is analyzed. The propagation equation of the error of rotary inertial navigation system is deduced, then, the modulation effect of single-axis rotation on various certainty errors of inertia measurement components is analyzed, moreover, the correctness of the theoretical analysis is verified by simulation. Through the analysis, it can be obtained that single-axis rotation can’t completely module the certainty errors of all inertial components; therefore, the precision of calibration scheme directly affects the positioning accuracy of the system. In order to improve the calibration precision of single-axis rotary fiber inertial components, a calibration scheme applied to the inertial measurement unit, the swivel table of which is not horizontal, is proposed. Mean while, the calibration model containing the error owing to the non-horizontal swivel table is established, then the error model of scale factor is changed into the form of polynomial function, rather than constant. According to the new established calibration model and calibration algorithm and considering the error characteristics of inertial components to diverge with time, the calibration process and scheme are designed. By doing experiments, the precision of the installation error and scale factor error of this new calibration scheme is verified to be improved by at least one more order of magnitude than that of the traditional one, and the bias calibration accuracy of inertial components is increased more than 2 times.Initial alignment in inertial navigation system is an essential step, and single-axis rotation modulation has no inhibition effect on initial alignment error, so the initial alignment error accumulates directly in navigation positioning error over time. In order to reduce the initial alignment error and make full use of the advantage of single-axis rotary inertial navigation system, a set of initial alignment scheme for single-axis rotary strap-down inertial navigation system is designed, which contains two positions of coarse alignment scheme and fine alignment scheme based on nondestructive Kalman filter. Due to the introduction of rotating mechanism in rotary inertial navigation system, the swivel table inversion error is also introduced in the process of initial alignment, in order to eliminate the error of swivel table, the nonlinear error state equation of initial alignment augmented by the inversion error is established, and the fine alignment scheme based on the UKF augmented by the inversion error is put forward as well. By doing hardware-in-loop simulations, the horizontal alignment precision of this new initial alignment scheme is verified to be improved by one more order of magnitude than that of the traditional one, and the position alignment precision is increased 2 times.Finally, a theoretical prototype of single-axis rotary fiber strap-down inertial navigation system is set up, and a large number of calibration and navigation experiment based on the built prototype verify the validity of the error identification method and compensation scheme. By doing the navigation experiment with the static base for 1 hour, the navigation positioning precision after stopping modulation is verified to be improved by one more order of magnitude than the traditional strap-down inertial navigation system. It provides the reference for the performance improvement of the fiber rotary strap-down inertial navigation system.