船载外测设备动态测量数据的误差分离  

Error Separation for Dynamic Measurement Data of Ship-borne Radiometric Equipment

作　　者：李晓勇[1] 张忠华[1] 朱伟康[1] 薛国虎[1] 李辉芬[1] 张同双[1] 

机构地区：[1]中国卫星海上测控部

出　　处：《飞行器测控学报》2009年第2期71-77,共7页Journal of Spacecraft TT&C Technology

摘　　要：为了全面分析研究测量船外测数据误差特性,须对测量数据中的误差进行分离,为此需找到合适的误差分离方法。通过对航天测量船船载测量设备工程背景复杂、数据产生过程复杂、误差源多等特点进行深入的分析,为船载设备测量数据建立回归分析模型,采用三次等距B样条函数作为测量数据误差分离的工具,用等距的标准B样条拟合测量系统的真实信号、系统误差,进而分离出随机误差。对大量原始测量数据误差的统计特性进行检验分析,得出了船载设备外测数据不同于陆基测量的一些误差特性——非独立性、子段非正态性、需用高阶AR模型描述。研究结果表明:使用标准等距B样条函数作为误差分离工具对于船载测量设备外测数据是适合的,样条拟合后的残差结果符合高阶AR模型。Overall analysis of the characteristics of ship-borne radiometric data requires a method for separation of errors from measurement data.Based on in-depth analysis of the complex engineering background,complicated data generation process and multiple error sources of the measurement systems of instrumentation ships,a regressive measurement data analysis model is established.Cubic B spline function is used as a tool to separate measurement data errors and standard equidistant B spline is applied to fit real signal and system error,from which random error is separated.Tests and analysis of the statistical attributes of a mass amount raw measurement data show that some error characteristics is different from data measured on land,such as independence and segmented non-normality,and this should be described by higher order auto-regressive(AR) model.The result shows that standard equidistant B spline function can be used to separate radiometric data errors and the residual result after spline fitting conforms to higher order AR model.

关 键 词：样条函数 数据拟合 航天测量船 外测数据 误差分离 残差检验 

分 类 号：V557[航空宇航科学与技术—人机与环境工程]