一种数据处理方法在初始对准中的应用

Application of a Method of Data Processing in Initial Alignment

阐述了基于扩展Kalman滤波的神经网络的模型设计过程，并将其运用到惯导系统的初始对准中；建立了惯导系统的非线性误差离散模型；并使用基于扩展Kalman滤波的神经网络的数据处理方法在初始对准中进行了仿真应用；仿真结果表明，采用该数据处理方法，不仅保证了对准的精度，而且提高了滤波的实时性。

The model of the Neural Network based on EKF is presented, which is applied to the alignment of inertial navigation systems （INS）, The nonlinear error discretization model of INS is derived. A method of data Processing of the Neural Network based on EKF is used in the initial alignment simulation, The simulation results show that the method of Data Processing ensures the accuracy of alignment and increases the real time of filtering.