Modeling and Inference of Resilient Systems

2017 Fiscal Year Final Research Report

• Project/Area Number: 26280092
• Research Category: Grant-in-Aid for Scientific Research (B)
• Allocation Type: Partial Multi-year Fund
• Section: 一般
• Research Field: Intelligent informatics
• Research Institution: National Institute of Informatics
• Principal Investigator: INOUE Katsumi 国立情報学研究所, 情報学プリンシプル研究系, 教授 (10252321)
• Co-Investigator(Kenkyū-buntansha): 沖本 天太 神戸大学, 海事科学研究科(研究院), 准教授 (10632432) ; Nicolas Schwind 国立研究開発法人産業技術総合研究所, その他部局等, 研究員 (60646397)
• Research Collaborator: RIBEIRO Tony エコールサントラルドゥナント, 博士研究員 ; MAGNIN Morgan エコールサントラルドゥナント, 准教授
• Project Period (FY): 2014-04-01 – 2017-03-31
• Keywords: 推論 / レジリエンス / 制約最適化 / 多目的最適化 / ロバスト性 / 安定性 / 平等性

Outline of Final Research Achievements

In this project, we formalize resilient systems that are tolerant to perturbations and are sustainable for maintaining their functionality, and consider several inference problems on this formalization, which can be used for designing resilient systems. Contributions are summarized as the three points: (A) The model for resilient systems, called SR model, has been extended and several parameterized properties have been improved to capture the resilience of systems. (B) By formalizing the resilience problem in dynamic systems as multi-objective constraint optimization problems, both exact algorithms to get the resilient solutions and approximate algorithms to compute only some representative Pareto solutions have been developed. (C) The notion of resilience has been applied to many applications, including team formation, robot rescue, biological oscillatory systems, belief revision of agents in social networks, then several computational properties have been analyzed for those systems.

Free Research Field

人工知能（推論，制約プログラミング，マルチエージェントシステム）