摘要
为进一步改正电离层延迟,在研究分析Klobuchar模型电离层模型的总电子含量误差时发现其存在着一些周期性规律信息。针对这些误差信息,提出了利用K-折交叉验证方法优化广义回归神经网络(generalized regression neural network,GRNN)径向学习速度,并建立基于Klobuchar模型的总电子含量误差补偿模型,对这些误差信息进行预测和补偿。试验结果表明,优化后的误差模型对不同地区和不同季节下电离层电子含量误差具有较好的预报精度和拟合效果。利用该模型对Klobuchar模型进行误差补偿,可将该模型总电子含量预报误差减小32%-90%,提高了改正精度。
In order to further correct the ionospheric delay,when studying and analyzing the total electron content error of Klobuchar model ionospheric model,it was found that there are some periodic law information.Aiming at this error information,this paper proposes a K-fold cross-validation method to optimize the radial learning speed of the generalized regression neural network(GRNN),and establish a total electron content error compensation model based on the Klobuchar model to predict and compensate these error information.The simulation results show that the model has good fitting ability and prediction effect on the error model of Klobuchar ionospheric total electron content in different regions and different seasons.Using this model to compensate the Klobuchar can reduce the error of vertical total electron content of the model by 32%-90%and improve the correction accuracy.
作者
简益梅
许承东
王倚文
彭雅奇
JIAN Yi-mei;XU Cheng-dong;WANG Yi-wen;PENG Ya-qi(School of Aerospace Engineering,Beijing Institute of Technology,Beijing 100081,China;Beijing Institute of Spacecraft System Engineering,Beijing 100094,China;China Helicopter Research and Development Institute,Tianjin Helicopter Research and Development Center,Tianjin 300000,China)
出处
《计算机仿真》
北大核心
2022年第8期45-50,共6页
Computer Simulation
关键词
电离层延迟
误差补偿
总电子含量
神经网络
Ionospheric delay
Error compensation
Vertical total electron content
Neural network