基于BP神经网络的简易地磁场矢量测量算法
A Simple Geomagnetic Field Vector Measurement Method Based on BP Neural Network
李小平 1陈瑞华 1李松 1霍鹏飞1
作者信息
- 1. 中国兵器工业集团第二一二研究所,陕西西安 710065
- 折叠
摘要
针对利用IGRF13模型在线解算地磁矢量存在的非线性以及计算实时性差的问题,提出一种基于BP神经网络的简易地磁场矢量测量算法.该方法通过BP神经网络建立起非线性的等效地磁场神经网络模型,并利用IGRF13模型数据对其进行训练,实际使用时仅需要测量载体经度、纬度以及海拔高度便可以通过简单加权计算得到当天测量点地磁场矢量,同时可以避免使用磁阻传感器测量地磁场存在参数标定以及易受周围环境干扰的问题.通过仿真验证,与在线解算算法相比,该算法计算量小,准确度满足要求,随着IGRF13模型的更新,可快速训练并建立更新后模型.
Abstract
Aiming at the problems of nonlinearity and poor real-time calculation of geomagnetic vector by using IGRF13 model online,this paper proposed a simple geomagnetic vector measurement method based on BP neural network.This method firstly established a nonlinear equivalent through BP neural network.Geomagnetic field neural network model,and used the IGRF13 model data to train it.In actual use,only the longitude,latitude and altitude of the carrier could be measured to obtain the geomagnetic field vector of the measurement point of the day through simple weighted calculation,and the use of magnetoresistive sensors could be avoided.The measurement of the geomagnetic field had the problems of parameter calibration and easy to be disturbed by the surrounding environment.Finally,the effectiveness of this method was verified through simulation.Compared with online solving algorithms,this algorithm had a smaller computational complexity and met the accuracy re-quirements.With the update of the IGRF13 model,it could quickly train and establish the updated model.
关键词
地磁测量/BP神经网络/地球磁场模型Key words
geomagnetic survey/BP neural network/earth magnetic field model引用本文复制引用
出版年
2024
NETL
NSTL
万方数据