残差最小化学習ニューラルネットワークを用いた電離圏トモグラフィーの性能評価
残差最小化学習ニューラルネットワークを用いた電離圏トモグラフィーの性能評価
カテゴリ: 論文誌(論文単位)
グループ名: 【A】基礎・材料・共通部門
発行日: 2015/02/01
タイトル(英語): Validation of Ionospheric Tomography using Residual Minimization Training Neural Network
著者名: 廣岡 伸治(千葉大学大学院理学研究科/國立中央大學太空科學研究所),服部 克巳(千葉大学大学院理学研究科)
著者名(英語): Shinji Hirooka (Graduate School of Science, Chiba University/Graduate Institute of Space Science, National Central University), Katsumi Hattori (Graduate School of Science, Chiba University)
キーワード: 残差最小化学習ニューラルネットワーク,電離圏トモグラフィー,性能評価 residual minimization training neural network,ionospheric tomography,validation
要約(英語): A numerical simulation has been done to evaluate the performance of the ionospheric tomography using the residual minimization training neural network (RMTNN) method. The results indicated that reconstruction with high-precision is possible when the standard deviation of the noise is about 2.5% or less of the average value of observed data (Slant TEC: STEC). Moreover, in the daytime when the value of STEC becomes large, the signal to noise ratio (SNR) increases and reconstruction accuracy becomes high; at night when the SNR falls conversely, it becomes low. Results of detectability tests show that the RMTNN method has a good performance around F-layer height with shape and peak intensity reconstruction. In conclusion, the developed RMTNN ionospheric tomography is effective in reconstructing 3D electron density distribution from realistic STEC data in the daytime, and is able to estimate images around F-layer.
本誌: 電気学会論文誌A(基礎・材料・共通部門誌) Vol.135 No.2 (2015) 特集:電気電子絶縁材料システムシンポジウム
本誌掲載ページ: 117-123 p
原稿種別: 論文/日本語
電子版へのリンク: https://www.jstage.jst.go.jp/article/ieejfms/135/2/135_117/_article/-char/ja/
受取状況を読み込めませんでした
