基于小样本数据的惯性平台系统贮存可靠度的预测方法  

Inertial Platform System Storage Reliability Prediction Method Based on Small Sample Data

作　　者：杨宝琨[1] 康兴无 

机构地区：[1]北京航天自动控制研究所,北京100854 [2]第二炮兵工程大学,西安710025

出　　处：《航天控制》2015年第5期90-95,共6页Aerospace Control

摘　　要：针对小样本的惯性平台系统贮存可靠性问题,开展了系统的方法研究。首先,分析了惯性平台系统定检数据的特点;其次,引入最小二乘支持向量机的预测方法对惯性平台系统贮存可靠度进行预测。数值计算结果表明,与多项式回归预测方法和灰色预测方法相比较,预测结果更加科学、准确。A method of the systematic approach is proposed,which is based on the problem of inertial platform system small sample storage reliability. Firstly,the characteristics of inertial platform system regular inspection data are analyzed. Secondly,the least squares support vector machine prediction method is used to predict the inertial platform system storage reliability. The numerical results show that the presented method has a more scientific and accurate prediction result that is compared to the polynomial regression method and gray method.

关 键 词：惯性平台系统 定期检测数据 小样本 贮存可靠性 支持向量机 

分 类 号：V240.2[航空宇航科学与技术—飞行器设计]