Signal processing on graphs with uncertainties

2023 Fiscal Year Final Research Report

• Project/Area Number: 22K21287
• Research Category: Grant-in-Aid for Research Activity Start-up
• Allocation Type: Multi-year Fund
• Review Section: 1001:Information science, computer engineering, and related fields
• Research Institution: Tokyo University of Agriculture and Technology (2023); Tokyo University of Science (2022)
• Principal Investigator: Yamada Koki 東京農工大学, 工学(系)研究科(研究院), 特任助教 (70965524)
• Project Period (FY): 2022-08-31 – 2024-03-31
• Keywords: グラフ信号処理 / グラフフィルタ / 電力系統 / 状態推定 / グラフ信号復元

Outline of Final Research Achievements

In this research project, we addressed signal processing methods in the presence of network uncertainties, such as power systems and transportation networks. We　investigated and developed methods for graph filter transition, time-varying graph signal restoration, and graph signal sampling, which are important issues in the field of graph signal processing. We applied the developed methods to state estimation of power systems and time-frequency analysis of acoustic signals, and demonstrated that they demonstrated superior performance compared to conventional methods.

Free Research Field

グラフ信号処理

Academic Significance and Societal Importance of the Research Achievements

確率密度比推定を利用し，あるネットワーク上で学習したグラフフィルタを異なるネットワーク上でも適用可能とする技術を開発した．本研究成果を電力システムの状態推定に適用し，電力網の情報を適切に取り入れることで，より高精度な推定を実現した．本研究の成果は，状態推定問題に限らず電力システム分野の重要課題であるセンサ配置問題やセキュリティリスク検知などにも応用可能な技術である．