摘要
短波信道由于受电离层的非线性变化影响而不能及时选到最佳频率,严重制约了短波通信系统的效能发挥。为了提高短波频率预测及选频的准确性,在总结前人关于短波频率预测经验的基础上,结合人工智能技术在非线性时间序列预测方面取得的成就,提出了一种短波通信频率的预测方法,该方法结合相空间重构技术和模糊小波神经网络技术,并在数据预处理阶段采用奇异值分解对历史数据进行降噪处理,实验结果表明,该方法比其他预测方法的精度有很大的提高。
Due to the non-linear variety of the ionosphere,the accurate forecast of HF radio communication frequency is very difficult.In this paper,a prediction method for HF radio communication frequency is presented.The method,in combination of phase space reconstruction and fuzzy wavelet neural network and by using singular value decomposition,implements noise-reduction processing of the historic data in data pretreatment.The experimental result indicates that this method is of fairly high precision as compared with other prediction methods.
出处
《通信技术》
2011年第4期37-39,共3页
Communications Technology
关键词
短波通信频率预测
相空间重构
模糊小波神经网络
奇异值分解
HF radio communication frequency prediction
phase space reconstruction
fuzzy wavelet neural network
singular value decomposition
