基于余弦基网络的弹载捷联惯导系统研究  

Application Research of Missile-borne SINS Based on Cosine Basis Function Neural Network

作　　者：查颖[1] 周长省[1] 李文琛[1] 

机构地区：[1]南京理工大学机械工程学院,南京210094

出　　处：《计算机测量与控制》2014年第4期1160-1162,共3页Computer Measurement &Control

摘　　要：捷联惯导系统是精确制导弹药中的一种高精度自主导航系统;针对目前弹载捷联惯性导航系统中陀螺仪精度不够理想的问题,建立了MEMS陀螺随机漂移信号的数学模型,并运用基于余弦基函数神经网络的数据融合技术对采集到的陀螺仪信号进行处理;实验结果表明,基于余弦基函数神经网络的数据融合技术可以有效地提高捷联惯导系统中MEMS陀螺信号处理的精度,融合后的信号精度平均提高38.45%。The strapdown inertial navigation system is a high precision autonomous navigation system in the precision--guided muni tions. The precision of the MEMS gyroscope in the missile-- borne strapdown inertial navigation system is not enough. According to this problem, a MEMS gyroscope random drift signal was established and the data fusion technology based on cosine basis function neural network was used to deal with the collected gyroscope signal. The experimental results show that, this method can effectively improve the MEMS gyro＇ s signal processing precision in the strapdown inertial navigation system and the signal processing accuracy has been increased by 38.45%.

关 键 词：捷联惯导系统 MEMS陀螺 数据融合 

分 类 号：TJ765.3[兵器科学与技术—武器系统与运用工程]