预测区间技术在航天器数据处理中的理论与应用研究  

Reserch on Prediction Interval Technology in Applying to Spacecraft data Processing

作　　者：安伟光[1] 孙振明 张辉 

机构地区：[1]哈尔滨工程大学航天工程系,哈尔滨150001 [2]济南黄河路桥工程中心,济南250032

出　　处：《宇航学报》2006年第B12期109-112,共4页Journal of Astronautics

基　　金：国防863高技术研究（2002AA721063）;国防十五预研项目（413200204）

摘　　要：面对航天领域收集到的大量遥测数据，对其整理、分析就需要运用统计的手段，采用了基于统计推断学的区间参数估计理论的预测区间技术，对传统的模型趋势预测研究进行了拓展。通过对航天数据及其预测模型的深入分析，把握其内在的统计规律，建立了模型预测与区间估计理论相结合的预测区间模型，给出未来数据在一定置信度下的变化区间。最后，将该技术应用于航天器数据预测实例分析，验证了该方法的有效性。Because it is necessary to apply statistical methods to collecting and analyzing mass telemetry data from space domain, this paper presents a prediction interval technology based on statistical interval estimation theory is studied to predict the space telemetry data and to develop trend model prediction. This paper analyzes spacecraft data and its predictive model and holds its inherent statistical law by the technology. Thus the paper builds the predictive interval model combining the interval estimation theory with the result of trend prediction, then produces a pair of bounds in which the further data will occur with a certain confidence degree. Finally, instantial analysis of predicting spacecraft data by this technology validates this method.

关 键 词：航天器 预测 预测区间 置信度 

分 类 号：TP206[自动化与计算机技术—检测技术与自动化装置] V47[自动化与计算机技术—控制科学与工程]