Implementation and dynamical analysis of a superconducting neural network solving a combinatorial optimization problem

• Project/Area Number: 24500271
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Sensitivity informatics/Soft computing
• Research Institution: Tohoku University
• Principal Investigator: ONOMI Takeshi 東北大学, 電気通信研究所, 助教 (70312676)
• Co-Investigator(Renkei-kenkyūsha): NAKAJIMA Koji 東北大学, 電気通信研究所, 教授 (60125622)
• Project Period (FY): 2012-04-01 – 2015-03-31
• Project Status: Completed (Fiscal Year 2014)
• Budget Amount: ¥5,070,000 (Direct Cost: ¥3,900,000、Indirect Cost: ¥1,170,000); Fiscal Year 2014: ¥1,560,000 (Direct Cost: ¥1,200,000、Indirect Cost: ¥360,000); Fiscal Year 2013: ¥1,560,000 (Direct Cost: ¥1,200,000、Indirect Cost: ¥360,000); Fiscal Year 2012: ¥1,950,000 (Direct Cost: ¥1,500,000、Indirect Cost: ¥450,000)
• Keywords: 感性情報学 / ニューラルネットワーク / 超伝導 / 組合せ最適化問題

Outline of Final Research Achievements

A Hopfield-type neural network solving a combinatorial optimization problem was designed by neuron circuits using superconducting devices. The dynamical analysis of the network was investigated by numerical simulations and operations of integrated circuits. A new neuron circuit with high current gain and flat output characteristics was developed and confirmed experimentally. The improvement of the network performance was confirmed by numerical simulations. On the other hand, a new relaxation oscillator using the new neural circuit was proposed, and its operation was confirmed numerically.

Research Products

• [Journal Article] An improved superconducting neural circuit and its application for a neural network solving a combinatorial optimization problem (2014)
  • Author(s): T. Onomi
  • Journal Title: Journal of Physics: Conference Series Volume: 507 Issue: 4 Pages: 042029-042029
  • DOI: 10.1088/1742-6596/507/4/042029
  • Peer Reviewed / Open Access / Acknowledgement Compliant
• [Journal Article] Neuron circuit using coupled SQUIDs gate with flat output characteristics for superconducting neural network (2014)
  • Author(s): Takeshi Onomi and Koji Nakajima
  • Journal Title: IEICE Transacsions on Electronics Volume: E97-C Pages: 173-177
  • NAID: 130003394704
  • Peer Reviewed
• [Presentation] 配線包囲型構造による超伝導マイクロストリップライン間の磁気結合度の評価 (2015)
  • Author(s): 小野美 武、中島 康治
  • Organizer: 第６２回応用物理学会春季学術講演会
  • Place of Presentation: 東海大学（平塚市）
  • Year and Date: 2015-03-11 – 2015-03-14
• [Presentation] Relaxation Oscillator Using Superconducting Schmitt Trigger Inverter (2014)
  • Author(s): T. Onomi
  • Organizer: 2014 International Symposium on Nonlinear Theory and Its Applications
  • Place of Presentation: Luzern, Switzerland
  • Year and Date: 2014-09-14 – 2014-09-18
• [Presentation] Back Propagation Learning Based on an IDL Model (2014)
  • Author(s): Y. Horiuchi, Y. Hayakawa, T. Onomi, and K. Nakajima
  • Organizer: 2014 International Symposium on Nonlinear Theory and Its Applications
  • Place of Presentation: Luzern, Switzerland
  • Year and Date: 2014-09-14 – 2014-09-18
• [Presentation] Performance analysis of Bidirectional Associative Memories by using the Inverse Function Delayless model (2014)
  • Author(s): C. Bao, T. Onomi, Y. Hayakawa, S. Sato, and K. Nakajima
  • Organizer: 2014 International Symposium on Nonlinear Theory and Its Applications
  • Place of Presentation: Luzern, Switzerland
  • Year and Date: 2014-09-14 – 2014-09-18
• [Presentation] 超伝導シュミットトリガーインバータとその応用 (2014)
  • Author(s): 小野美 武
  • Organizer: 電子情報通信学会超伝導エレクトロニクス研究会
  • Place of Presentation: 機械振興会館（東京都）
  • Year and Date: 2014-07-23
• [Presentation] High-speed single flux quantum parallel multiplier using Dadda type partial product addition (2014)
  • Author(s): A. Yamada, T. Onomi, and K. Nakajima
  • Organizer: 7th International WorkShop on New Group IV Semiconductor Nanoelectronics and JSPS Core-to-Core Program Joint Seminar, "Atomically Controlled Processing for Ultralarge Scale Integration"
  • Place of Presentation: 東北大学（仙台市）
• [Presentation] 平坦な入出力特性を持つ超伝導しきい値素子の設計と試作 (2014)
  • Author(s): 小野美武、中島康治
  • Organizer: 第61回応用物理学会春季学術講演会
  • Place of Presentation: 青山学院大学（相模原市）
• [Presentation] 単一磁束量子対を伝送するrf-SQUIDラダー回路の動作特性 (2014)
  • Author(s): 辻裕也、小野美武、中島康治
  • Organizer: 2014年電子情報通信学会総合大会
  • Place of Presentation: 新潟大学（新潟市）
• [Presentation] An improved superconducting neural circuit and its application for a neural network solving a combinatorial optimization problem (2013)
  • Author(s): Takeshi Onomi
  • Organizer: 11th European Conference on Applied Superconductivity
  • Place of Presentation: Genova, Italy
• [Presentation] SFQ高速並列乗算器における各種構成回路の比較 (2013)
  • Author(s): 山田朗文、小野美武、中島康治
  • Organizer: 電子情報通信学会2013年ソサエティ大会
  • Place of Presentation: 福岡工業大学（福岡市）
• [Presentation] Characteristics of rf-SQUID Ladder Circuits (2013)
  • Author(s): Y. Tsuji, T. Onomi, and K. Nakajima
  • Organizer: Superconducting SFQ VLSI Workshop SSV 2013
  • Place of Presentation: 産業技術総合研究所（つくば市）
• [Presentation] Comparative Study of SFQ Parallel Multipliers (2013)
  • Author(s): A. Yamada, T. Onomi, and K. Nakajima
  • Organizer: Superconducting SFQ VLSI Workshop SSV 2013
  • Place of Presentation: 産業技術総合研究所（つくば市）
• [Presentation] SFQ高速並列乗算器における部分積加算回路の比較 (2013)
  • Author(s): 山田朗文、小野美武、佐藤茂雄、中島康治
  • Organizer: 2013年電子情報通信学会総合大会
  • Place of Presentation: 岐阜大学（岐阜市）
• [Presentation] ｒｆ-SQUIDラダー回路の特性解析 (2013)
  • Author(s): 辻裕也、小野美武、佐藤茂雄、中島康治
  • Organizer: 2013年電子情報通信学会総合大会
  • Place of Presentation: 岐阜大学（岐阜市）
• [Presentation] 平坦な入出力特性を持つ超伝導ニューロン回路の設計と試作 (2013)
  • Author(s): 小野美武、中島康治
  • Organizer: 第60回応用物理学会春季学術講演会
  • Place of Presentation: 神奈川工科大学（厚木市）
• [Presentation] Design and Fabrication of an Improved Neural Circuit for Superconducting Neural Network Solving a Combinatorial Optimization Problem (2012)
  • Author(s): T. Onomi and K. Nakajima
  • Organizer: Superconducting SFQ VLSI Workshop 2012
  • Place of Presentation: 名古屋大学（名古屋市）