一种采用小波神经网络的GPS精密单点定位方法 |
| |
引用本文: | 米洋,陈家斌,刘红光,王秋帆,傅金琳. 一种采用小波神经网络的GPS精密单点定位方法[J]. 中国惯性技术学报, 2016, 0(3): 337-341. DOI: 10.13695/j.cnki.12-1222/o3.2016.03.011 |
| |
作者姓名: | 米洋 陈家斌 刘红光 王秋帆 傅金琳 |
| |
作者单位: | 1. 北京理工大学自动化学院,北京,100081;2. 天津航海仪器研究所,天津,300131;3. 北京自动化控制设备研究所,北京,100074 |
| |
基金项目: | 船舶预研支撑技术基金项目(14JZ3.9.2) |
| |
摘 要: | 针对GPS精密单点定位对高精度的需求,提出了一种采用小波神经网络的GPS精密单点定位解算方法。该方法利用小波变换和神经网络学习功能,无需准确系统先验信息,误差函数能够快速收敛,逼近真实误差模型,从而提高GPS精密单点定位精度。仿真结果表明,静态条件下与传统最小二乘法和卡尔曼滤波算法相比,该算法定位收敛时间缩短50%,定位精度分别提升90%和50%。动态情况下,较最小二乘法和卡尔曼滤波算法定位精度提高20%~80%。
|
关 键 词: | GPS精密单点定位 小波变换 神经网络 收敛时间 |
GPS precise point positioning method using wavelet neural network |
| |
Abstract: | Aiming at the high precision demand for GPS precise point positioning, a GPS precise point positioning method using wavelet neural network is proposed. The method adopts wavelet transform and neural network learning function, and can make the error function rapidly convergent without needing accurate system priori information and can approximate the true error model, which improve the GPS precise point positioning accuracy. Simulation results show that the proposed algorithm can shorten the GPS precise point positioning time by more than 50% and improve positioning accuracy by 90% and 50% respectively compared to the traditional least square method and the Kalman filter algorithm under static conditions. Under dynamic conditions, the proposed algorithm can improve positioning accuracy by 20%~80% compared to the traditional least square method and the Kalman filter algorithm. |
| |
Keywords: | GPS precise point positioning wavelet transform neural network convergence time |
本文献已被 CNKI 万方数据 等数据库收录! |
|