基于人工神经网络的东亚电离层临界频率f_(o)F2长期变化趋势  

Long-term variation trend of east asian ionospheric critical frequency f_(o)F2 based on artificial neural network

作　　者：朱正平[1] 邓杰 ZHU Zhengping;DENG Jie(College of Electronics and Information Engineering,South-Central Minzu University,Wuhan 430074,China)

机构地区：[1]中南民族大学电子信息工程学院,武汉430074

出　　处：《中南民族大学学报（自然科学版）》2024年第6期753-758,共6页Journal of South-Central Minzu University(Natural Science Edition)

基　　金：国家自然科学基金资助项目(41474135)。

摘　　要：利用人工神经网络对东亚中纬地区电离层台站观测的F2层临界频率f_(o)F2进行长期趋势研究.用F107、Ap、地方时(Local Time,LT)、月份(Month)作为输入神经元,分别表示太阳活动、地磁活动、日变化和季节变化;用f_(o)F2的月中值作为输出神经元,通过训练网络获取f_(o)F2预测值,对预测值和观测值进行计算和处理,得到东亚中纬地区f_(o)F2长期变化趋势.结果表明:人工神经网络的方法较常用的回归方法能更有效地消除地磁活动对f_(o)F2的影响;这些站点的f_(o)F2随着年份的增长存在明显的长期负趋势,无明显的日变化和统一季节变化性.这对于全球电离层结构和运动变化规律,全球电离层经验模型构建和同化以及电离层特征参数和结构预测具有重要意义.The trend of F2 layer critical frequency f_(o)F2 of ionospheric stations in East Asia mid-latitude is analyzed using artificial neural network method.F107,Ap,Local Time(LT),Month is used as input neurons to represent solar activity,geomagnetic activity,diurnal and seasonal changes respectively.The monthly median value of f_(o)F2 is used as output neuron,and the predicted value of f_(o)F2 is obtained by training the network.The predicted and observed values are processed and calculated to obtain the long-term variation trend of f_(o)F2 in East Asia mid-latitude.The results show that the artificial neural network method can more effectively eliminate the influence of geomagnetic activity on f_(o)F2 than the commonly used regression method.There is a clear long-term negative trend in the f_(o)F2 of these sites with the increase of the year.And there is no obvious diurnal variation and uniform seasonal variability.These are of great significance for the global ionospheric structure and movement change law,the construction and assimilation of the global ionospheric empirical model,and the ionospheric characteristic parameters and structure prediction.

关 键 词：人工神经网络 电离层 F2层临界频率 太阳和地磁活动 

分 类 号：P352[天文地球—空间物理学]