| 弹道跟踪数据野值剔除方法性能分析 |
| |
| 引用本文： | 侯博文,王炯琦,周萱影,李冬,何章鸣.弹道跟踪数据野值剔除方法性能分析[J].上海航天,2018(4):91-100. |
| |
| 作者姓名： | 侯博文 王炯琦 周萱影 李冬 何章鸣 |
| |
| 作者单位： | 国防科技大学理学院;北京控制工程研究所;中国人民解放军91550部队94分队 |
| |
| 基金项目： | 国家自然科学基金(61773021,7) |
| |
| 摘 要： | 弹道跟踪数据中的野值会影响导弹目标定位精度。对用于弹道跟踪数据的野值剔除方法进行了综述。结合事后和实时两大类处理模式,分析了各野值剔除方法在理论和工程应用上的优缺点,给出了实际应用时的限制因素,建立了野值判决准则和相应的野值剔除实现步骤。仿真结果表明:野值剔除方法可以有效提高弹道跟踪数据的精度。事后处理模式中,基于格拉布斯准则的野值剔除效果最佳;实时处理模式中,基于自适应门限的五点线性预报法野值剔除效果相对较好,且时间复杂度相对较低,适合实时处理。
|
| 关 键 词： | 弹道跟踪数据 野值剔除 事后处理 实时处理 判决准则 精度 时间复杂度 剔除效果 |
| 收稿时间： | 2017/10/13 0:00:00 |
| 修稿时间： | 2018/1/8 0:00:00 |
Analysis on Performance of Ballistic Tracking Data Outlier Elimination Methods |
| |
| HOU Bowen,WANG Jiongqi,ZHOU Xuanying,LI Dong and HE Zhangming.Analysis on Performance of Ballistic Tracking Data Outlier Elimination Methods[J].Aerospace Shanghai,2018(4):91-100. |
| |
| Authors: | HOU Bowen WANG Jiongqi ZHOU Xuanying LI Dong and HE Zhangming |
| |
| Affiliation: | College of Science, National University of Defense Technology, Changsha 410072, Hunan, China,College of Science, National University of Defense Technology, Changsha 410072, Hunan, China; Beijing Institute of Control Engineering, Beijing 100086, China,College of Science, National University of Defense Technology, Changsha 410072, Hunan, China,Unit 94, PLA 91550, Dalian 116023, Liaoning, China and College of Science, National University of Defense Technology, Changsha 410072, Hunan, China; Beijing Institute of Control Engineering, Beijing 100086, China |
| |
| Abstract: | Outliers of ballistic tracking data can influence the positioning accuracy of missile targets. This paper summarizes the common outlier eliminating methods of ballistic tracking data. Combined with the off-line and on-line processing modes, we analyze advantages and disadvantages of the methods in terms of theory and practical application, and present the limiting factors in practical application. Based on the methods, the outlier elimination procedure and criterion are established. Simulation results indicate that outlier elimination methods can reduce the outliers in measurement data. In off-line data processing, the outlier elimination effect based on Grubbs criterion is the best. In on-line data processing, the five-point linear prediction data detection method based on self-adaptive threshold has the best effect. Its time complexity is much lower, which is suitable for real-time processing. |
| |
| Keywords: | ballistic tracking data outlier elimination off-line data processing on-line data processing judging criterion accuracy time complexity elimination effect |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《上海航天》浏览原始摘要信息 |
|
点击此处可从《上海航天》下载全文 |
|
