基于改进Elman神经网络的SINS动基座传递对准研究

Improved Elman Neural Network for SINS Transfer Alignment on the Moving Base

在运用Kalman滤波进行SINS动基座传递对准时,当模型存在误差或系统噪声不能反映实际噪声时,会降低滤波精度甚至导致滤波发散。针对这个问题,提出基于改进Elman神经网络的SINS动基座传递对准方法。首先通过增加输出层节点的反馈来改进普通的Elman神经网络模型,其次采用强跟踪滤波器对改进Elman神经网络进行训练。利用仿真数据对该算法进行验证,结果表明,该算法能够克服Kalman滤波的缺陷,提高传递对准精度达100%~150%。

In SINS transfer alignment on the moving base using Kalman filtering,when a inaccurate model is used or the systematic covariance matrix doesn t indicate the actual noise,it will degrade the filtering accuracy or even lead it to divergence.In order to solve this problem,a method based on improved Elman neural network was presented.First,the ordinary Elman neural network model was improved by adding a feedback on output layer.Then,the strong tracking filter was applied to train the improved Elman neural netw...