基于支撑矢量机的导弹惯性器件故障预报系统研究  

Missile Inertia Device Fault Prediction System Based on Support Vector Machine

作　　者：陈伟[1] 胡昌华[1] 方华元[1] 丁力[1] 

机构地区：[1]第二炮兵工程学院,西安710025

出　　处：《弹箭与制导学报》2005年第SB期506-509,共4页Journal of Projectiles,Rockets,Missiles and Guidance

摘　　要：文中针对导弹惯性器件故障预报系统的设计要求,考虑到支撑矢量机用于故障预报的优点,在导弹惯性器件的故障预报过程中应用了支撑矢量机的训练算法。把支撑矢量机、预测理论和故障诊断系统有机结合起来,建立了一个故障预报系统。利用该系统选取支撑矢量机预测模型对某导弹陀螺随机漂移进行建模和分析,实现了故障预报,实例预测结果证明,给出的支撑矢量机预测模型和训练算法是有效和可行的。According to design requirement of missile inertia device fault prediction system and the advantage of support vector machine(SVM).the training algorithm of SVM in missile inertia device fault prediction has been put forward.A missile inertia device fault prediction system has been established with SVM,fault prediction technique and fault diagno- sis system.This system with SVM forecasting model has been used to build model and analysis gyroscope drift of a cer- tain missile.The simulated experiment proved the forecasting model and the algorithm is valid and acceptable.

关 键 词：支撑矢量机 导弹惯性器件 故障预报 陀螺漂移 

分 类 号：E927[军事—军事装备学]