摘要
研究了带内部反馈的 Elman 神经网络的特性,提出了应用 Elman 神经网络的 INS/GPS 组合导航滤波计算方法,仿真表明神经网络滤波可以达到与卡尔曼滤波器相当的估计精度,对模型误差和噪声特性的变化具有较强的鲁棒性。
In this paper,the character of Elman neural network with inner feedback is studied,then an INS/GPS integrated navigation filtering algorithm using the neural network is proposed.Simulation results indicate that the algorithm,compared with Kalman filter,has similar accuracy and better rohustness to model uncertainties and noise characteristics variation.
出处
《弹箭与制导学报》
CSCD
北大核心
2003年第S2期91-94,共4页
Journal of Projectiles,Rockets,Missiles and Guidance
