摘要
短波通信频率是随着电离层变化的非线性时间序列,利用长短期记忆人工神经网络(LSTM)模型对时间序列的非线性关系处理的突出优势,将短波远程通信中两点间的可用频率作为单特征输入,通过时间序列的非线性运算预测未来的可用频率。介绍几种时间序列算法,在Python平台上的建模仿真,得出相应算法的预测结果,并对预测结果进行对比分析,最终得出运行时间较快,效果最好的长短期记忆人工神经网络LSTM模型更适用于两点间短波最高可用频率的预测,对远程短波通信保障具有重要意义。
HF frequency is a non-linear time series varying with the ionosphere.The long short-term memory artificial neural network model is used to deal with the non-linear relationship of time series.The available frequency of HF is used as a single char acteristic input to predict available frequencies through the nonlinear time series operation of funture.Several time series algorithms are introduced,and the simulation results are obtained through the Python platform.The prediction results of the corresponding algo rithms are obtained,and the prediction results are compared and analyzed.The long short-term memory artificial neural network model with the faster running time and the best effect is more suitable for the prediction of the maximum usable frequency of HF com munication between two points.
作者
尚教凯
张海勇
徐池
徐铭
SHANG Jiaokai;ZHANG Haiyong;XU Chi;XU Ming(Department of Information Systems,Dalian Navy Academy,Dalian 116018;Dalian Changxing Island Economic Zone Primary and Secondary School Quality Education Practice Base,Dalian 116317)
出处
《舰船电子工程》
2019年第11期76-78,88,共4页
Ship Electronic Engineering
基金
国家社会科学基金项目（编号：15GJ003-208）资助
