摘要
利用多输入多输出小波神经网络具有结构简单、计算量少的优点,将其应用于静基座捷联惯导系统的初始对准.并利用北向、东向失准角快速估计地向陀螺的失准角,得到小波神经网络的训练样本.对小波神经网络采用随机梯度法进行训练.仿真结果表明该方法收敛速度快,能满足惯导系统实时性的要求.
The multipleinputmultipleoutput (MIMO) wavelet network is applied to initial alignment of Strapdown Inertial Navigation System (SINS) for stationary base because of its easy structure and less calculation.A rapid estimation algorithm is also provided to obtain the training samples for the wavelet network.Using the stochastic gradient algorithm,the parameters of the wavelet network can be adjusted.The simulation results show that this method is fast convergent.So it can meet the needs of real time situation.
出处
《内蒙古大学学报（自然科学版）》
CAS
CSCD
北大核心
2003年第4期468-471,共4页
Journal of Inner Mongolia University：Natural Science Edition
关键词
小波神经网络
初始对准
捷联
静基座
wavelet networks
initial alignment
strapdown
stationary base
