第8回

【第8回】富士通次世代コンピューティングセミナー

• 日時
  • 2024年8月28日（水）15:00-17:00

• 講演
  1. “Progress on highly parallel ensemble solvers”
     • Yuji Shinano (Zuse Institute Berlin)

       平成９年３月東京理科大博士課程修了．平成９年４月より東京理科大助手．東京農工大准教授を経て，平成２２年４月より Zuse Institute Berlin (ZIB) 研究員

     • Abstract
       • The Ubiquity Generator (UG) Framework was initially developed to parallelize advanced branch-and-bound based solvers. With the introduction of version 1.0, UG transitioned into a high-level task parallelization framework, enabling the parallelization of a broad range of state-of-the-art solvers. This framework facilitates the integration of multiple algorithmic implementations into a parallel solver. This presentation will highlight several success stories where UG has been utilized to solve previously intractable instances of Mixed Integer Programming, Steiner Tree, Quadratic Assignment, and Quadratic Unconstrained Binary Problems, as well as to set new benchmarks for Shortest Vector Problem instances. Finally, the most recent developments in the UG Framework will be discussed.
  2. “Role of Hydrodynamics Stability Analysis in Energy Transition”
     • Thi Thai LE (Zuse Institute Berlin)

       Dr. LE is the Head of the Research Group of Predictive Methods at the Applied Algorithmic Intelligence Methods (A²IM) Department at the Zuse Institute Berlin. Her primary research focus centers around numerical methods and computations in Fluid Dynamics. An outstanding application of her research lies in supporting the right decision to optimize operations of renewable energy systems, including hydrogen and wind power. In 2019, Dr. Le earned her Doctorate in Functional Mathematics at Kyushu University in Japan, financially supported by the MEXT Scholarship provided by the Japan Ministry of Education, Culture, Sports, Science, and Technology. Afterward, she actively engaged in research systems in Germany, collaborating with renowned institutions such as the Applied University of Sciences Nuremberg, the Technical University Berlin, and the Zuse Institute Berlin.

     • Abstract
       • Hydrodynamics instability results from complex interactions between fluid forces and structural dynamics, giving rise to unpredictable flow patterns and dynamic fluctuations. This instability becomes a crucial factor in comprehending and optimizing the operations of various industrial technologies. This talk will provide an overview of how hydrodynamic instability can support maximizing the extraction and injection of hydrogen from/into underground storage efficiency, a critical solution in the energy transition to cleaner and more effective power sources in Germany and globally.
  3. “The process of developing vehicle controllers for handling evasive maneuvers”
     • Amauri da Silva Junior (Technische Hochschule Ingolstadt and Royal Melbourne Institute of Technology)

       Amauri da Silva Junior obtained his diploma in automotive engineering from the Federal University of Santa Catarina (UFSC), in Brazil. He is currently pursuing a Ph.D. at the Royal Melbourne Institute of Technology (RMIT) University in Melbourne, Australia, while also working as a researcher at Technische Hochschule Ingolstadt (THI) in Germany. His research interests include vehicle control for autonomous vehicles and the development of emergency maneuvering controllers.

     • Abstract
       • With the advancement of autonomous vehicle technology, it is essential to design controllers capable of handling evasive maneuvers to ensure the safety of both the vehicle and road users. This talk covers the comprehensive process of developing and implementing vehicle controllers specially designed for emergencies. We will talk about the main pillars for the development of autonomous vehicles, as well as the classification of active safety and passive safety. The importance of controllers for emergencies is highlighted in real-life situations. We will discuss how to design a controller from an idea to the point where the controller is tested in a vehicle. The main methods used for controllers are presented, as well as the controller developed in this project. We will discuss how to use classic controller models with the use of artificial intelligence, as well as the benefits and drawbacks of each approach. A short discussion on the importance and development of scaled vehicles is presented, as well as a comparison of simulation with real test results.
• 場所（ハイブリッド開催）
  • オフライン：広島大学東広島キャンパス 情報メディア教育研究センター本館2階セミナー室1, 2, 3
    • 東広島キャンパスマップ
  • オンライン：Microsoft Teams会議
• 参加方法
  • こちらの参加登録フォームよりお申込みください
  • 学内からの参加は基本的に登録不要です（オンライン参加でURLがわからない場合はご登録ください）

現地30名＋オンライン20名の約50名に参加していただきました