一种神经网络实现GPS导航定位算法

A Neural Network GPS Navigation Algorithm

研究ＧＰＳ接收机状态解算的估计方法，方法利用Ｈｏｐｆｉｅｌｄ神经网络对ＧＰＳ接收机数据进行处理．结果以统计算术平均误差和均方根误差作为衡量性能的指标，神经网络方法与迭代最小均方误差算法相比较定位误差减小．结论利用神经网络实现ＧＰＳ接收机状态解算是可行的，且与迭代最小均方误差算法比较，可提高定位精度．

Aim study the estimation algorithm of the GPS receiver state solution.Methods A Hopfield neural network was used to process the GPS receiver data.Results Compared with the recursive least square(RMS) algorithm,the solution error of Neural Network was reduced both in mean bias and in rms error. Conclusion It is a success to use neural network in the estimation of GPS navigation solution, and THE accuracy is improved.