一种星载GNSS-R海风反演的卡尔曼滤波模型  

Kalman filter model for spaceborne GNSS-R ocean wind retrieval

作　　者：李中奎 张波[1] 杨东凯[1] 张国栋 LI Zhongkui;ZHANG Bo;YANG Dongkai;ZHANG Guodong(School of Electronics and Information Engineering,Beihangjing University,Beijing 100191,China)

机构地区：[1]北京航空航天大学电子与信息工程学院,北京100191

出　　处：《导航定位学报》2020年第4期31-38,共8页Journal of Navigation and Positioning

基　　金：国家重点研发计划项目(2017YFB0502802)。

摘　　要：针对传统星载全球卫星导航系统反射计(GNSS-R)测风,通过建立经验性的地球物理模型函数(GMF)进行风速反演时,会因反射信号数据质量较差而出现异常结果,若执行严格的数据质量控制,则会降低数据利用率,最终影响星载GNSS-R测风的空间覆盖率的问题,提出1种基于卡尔曼(Kalman)滤波的风速反演算法:利用差分整合移动平均自回归模型(ARIMA)得到状态方程和观测方程,并将GMF反演的风速值作为观测值,从而建立Kalman滤波模型,实现星载GNSS-R风速反演的实时校正和优化。实验结果表明,该方法的风速反演均方根误差能够满足风速测量要求,并有效提升星载GNSS-R测风的空间覆盖率。Aiming at the problems that it is liable to poor quality of reflected signal data leading to abnormal results for the traditional spaceborne GNSS-reflectometry wind measurement in the wind speed retrieval by establishing an empirical geophysical model function(GMF),and it is easily to reduce the data utilization and ultimately affect the spatial coverage of the spaceborne GNSS-R wind measurement if strict data quality control is performed,the paper proposed a wind speed retrieval algorithm based on Kalman filtering:the state equation and the observation equation were obtained by ARIMA,and the wind speed value of the GMF retrieval was taken as the observation value,then the Kalman filter model was established to realize the real-time correction and optimization of spaceborne GNSS-R wind speed inversion.Experimental result showed that the root mean square error of the wind speed retrieval with the proposed method could meet the requirement of wind speed measurment,and efficiently increase the spatial coverage of spaceborne GNSS-R wind measurements.

关 键 词：星载全球卫星导航系统反射计 风速反演 卡尔曼滤波 地球物理模型函数经验模型 差分整合移动平均自回归模型 

分 类 号：P228[天文地球—大地测量学与测量工程]