| 动态系统的抗差Kaliman滤波 |
| |
| 引用本文： | 杨元喜.动态系统的抗差Kaliman滤波[J].测绘学院学报,1997(2). |
| |
| 作者姓名： | 杨元喜 |
| |
| 作者单位： | 解放军测绘学院!郑州，450052 |
| |
| 基金项目： | 国家自然科学基金，德国Humboldt基金 |
| |
| 摘 要： | 离散历元的动态观测量及其相应的动态模型可能存在异常，若数据处理模型不考虑对这些异常的特别处理，则动态模型参数估值及其所提供的动态信息将极不可靠。基于贝叶斯统计和抗差估计原理，我们构造了一种抗差滤波算法。该算法考虑观测分布和参数验前分布均为污染分布。并利用一个实测网验算该算法和模型的可靠性。
|
| 关 键 词： | 动态模型 异常观测 贝叶斯分布 抗差滤波 |
Robust Kalman Filter for Dynamic Systems |
| |
| Yang Yuanxi.Robust Kalman Filter for Dynamic Systems[J].Journal of Institute of Surveying and Mapping,1997(2). |
| |
| Authors: | Yang Yuanxi |
| |
| Institution: | Yang Yuanxi |
| |
| Abstract: | Dynamic measurements at discrete epochs and the corresponding dynarnic system models may exist outliers which may have dramatically effects on the state parameter estimates lf the estimation method does not consider to control these outliers. Based on Bayesian statistics and the robust estimation principle. a general robust Kalman filter is constructed in which the contaminated distributions for both of the measurements and the predicted state parameters are consldered. A set of special estimators relevant to particular assumptions are given. A practical dynamic geodetic network is adjusted as an example. For the analysis of error influence. the error influence functions and the empirical influence functions corresponding to the robust filters and the Kalman filter are derived. |
| |
| Keywords: | Dynamic model Outliers Bayesian distributionl Robust filter |
| 本文献已被 CNKI 等数据库收录！ |
