摘要
研究了一种卡尔曼滤波算法 ,通过算例对我国 1 998年洪水期某大坝的实测GPS动态变形数据 ,采用自编软件 (用C语言 )进行处理 ,其成果图形象直观地显示出滤波值数据曲线与原始数据曲线的变化趋势基本一致 ,同时验前与验后的单位权方差σ∧20 很相近 ,并且滤波值中误差σ∧滤 也比观测值中误差σ∧测 小得多 .该离散卡尔曼滤波数学模型较好地模拟了目标系统的物理变化规律 ,其程序运算结果对于改善GPS动态变形的数据精度的效果比较理想 .
An approach of calculation of Kalman filtering is studied.The program adapted from the computer language C was applied to process the GPS kinematic deformation data of actual measurements for some big dam in our country in the 1998_flood season as a testing case.It is obvious that the achievement chart of this example vividly indicates the changing tendency of the data curve of filtering value very close to the one of raw data.Meanwhile,variance of unit weight before statistical test is approximately equal to the one after statistical test,there is less standard error of filtering values than the one of measured values.The mathematical model of this discrete Kalman filtering appropriately simulates the law of physical variation of observation system.It is ideal to improve the GPS data accuracy of kinematic deformation by carrying on the software of Kalman filtering.
出处
《桂林工学院学报》
2002年第3期234-238,共5页
Journal of Guilin University of Technology
基金
广西壮族自治区教育厅资助项目 (1998- 16 9)
