摘要
GPS导航解算中常用最小二乘算法。随着高动态用户需求精度的不断提高，且由于线性化忽略高次项，初始值精度低以及差分后剩余或放大误差的存在。导航解精度很难满足高动态用户的需求。为此，本文基于BP神经网络的非线性逼近性能。给出了基于BP神经网络的GPS导航算法。实测数据计算结果表明该算法能够真实地反映载体运动轨迹，其导航解的精度和可靠性有明显的提高。
LS algorithm is usually used estimate unknown parameters in GPS dynamic data processing. With the need of more precise state estimation,the reliability of the I.S estimated states in navigation is not sat isfied with precise kinematic GPS users ,for the high order terms in linearing model are ignored and the influence of the low precise initial value and the amplified errors in differential GPS system are existed. So a GPS Navigation Algorithm based on BP neural network is improved in this paper. It is shown, by calculations and analysis, that the GPS navigation algorithm based on BP neural network can give more actual and reliable parameter estimates of the maneuvering vehicles.
出处
《测绘科学与工程》
2006年第2期47-50,共4页
Geomatics Science and Engineering
基金
国家自然科学基金资助（40274002和40474001）
