航空矢量重力测量中光纤陀螺随机漂移误差实时补偿方法

发布时间：2018-08-24 20:53

【摘要】：光纤陀螺随机漂移误差是影响航空矢量重力测量系统姿态解算精度的关键因素。建立模型并在输出中对其补偿是抑制该项误差的有效方法。针对传统ARMA模型只能对平稳随机漂移误差建模,且模型无法满足实时滤波需求的问题,本文引入适用于非平稳随机漂移误差的ARIMA模型,同时给出详细的建模过程,并提出采用实时平均算法消除原始采样序列中常值分量的思路,实现了随机漂移误差的实时Kalman滤波估计。基于本文所提出的模型和实时滤波算法,对光纤陀螺实测数据进行分析,结果表明处理后信号中随机漂移误差的方差减小了46.5%。Allan方差分析结果表明,滤波后角度随机游走系数和角速率随机游走系数分别降低了约50%和40%。本文的结果说明ARIMA模型能够准确描述陀螺的非平稳随机漂移误差。基于实时平均算法的Kalman滤波可实现随机漂移误差的在线估计,有望提高航空矢量重力测量系统的姿态解算精度。
[Abstract]:Random drift error of fiber optic gyroscope (fog) is a key factor to affect attitude resolution accuracy of airborne vector gravity measurement system. Establishing the model and compensating it in the output is an effective method to restrain the error. Because the traditional ARMA model can only model the stationary random drift error, and the model can not meet the requirement of real-time filtering, this paper introduces the ARIMA model which is suitable for the non-stationary random drift error, and gives the detailed modeling process. A real-time averaging algorithm is proposed to eliminate the constant component in the original sampling sequence, and the real-time Kalman filter estimation of the random drift error is realized. Based on the proposed model and real-time filtering algorithm, the measured data of fog are analyzed. The results show that the variance of random drift error in the processed signal reduces the variance of 46.5%.Allan variance analysis. After filtering, the angle random walk coefficient and angular rate random walk coefficient are reduced by about 50% and 40% respectively. The results show that the ARIMA model can accurately describe the non-stationary random drift error of gyroscope. The Kalman filter based on real-time averaging algorithm can realize the on-line estimation of random drift error, which is expected to improve the attitude resolution accuracy of airborne vector gravimetric measurement system.
【作者单位】：武汉大学测绘学院;武汉大学地球空间环境与大地测量教育部重点实验室;
【基金】：国家自然科学基金(41210006;41504016) 国家973计划(2013CB733301) 地球空间环境与大地测量教育部重点实验室测绘基础研究基金(14-02-01)~~
【分类号】：P223.4

【相似文献】