动态新息GM(1,1)在卫星电池阵功率预测中的应用  

Application of dynamic innovation GM(1,1) to power prediction of satellite battery array

作　　者：谷吉海[1] 姜兴渭[1] 巴兴强[2] 范显峰[1] 黄文虎[1] 

机构地区：[1]哈尔滨工业大学航天学院,黑龙江哈尔滨150001 [2]东北林业大学,黑龙江哈尔滨150040

出　　处：《哈尔滨工业大学学报》2003年第1期28-30,34,共4页Journal of Harbin Institute of Technology

基　　金：国家高技术研究发展计划资助项目 ( 863 2 0 0 2AA72 10 63 )

摘　　要：根据观测到的历史数据对卫星电池阵输出功率的未来值进行预测 ,可实现对电池阵性能劣化的早期预报 .灰色预测方法的GM(1,1)模型只适合对单调递增或递减时间序列进行预测 ,针对卫星电池阵输出功率具有波动变化的趋势特征 ,采用实时在线的方法 ,建立了动态新息GM (1,1)预测模型 ,经实例预测验证 ,动态新息GM(1,1)模型可明显地提高预测精度 ,并且能对电池阵输出功率的波动变化趋势正确预测 .The early forecast of satellite battery array's performance attenuation can be realized according to the prediction of historical datum of satellite battery array's output power. The GM (1,1) of grey prediction method fits model for time-series prediction of monotonic increasing or digression only. A real-time method is proposed according to the fluctuant character of batter matrix output power. The prediction model of dynamic innovation GM(1,1) is founded. The prediction precision is improved greatly by examples. The fluctuant tendency of batter array output power can be predicted exactly. The method of model making has referential value for other prediction of non-monotonic time-series.

关 键 词：灰色预测 动态新息GM(1 1)模型 卫星电池阵 功率预测 输出功率 卫星电源系统 

分 类 号：V442[航空宇航科学与技术—飞行器设计] V474