デジタル打音検査とAI・シミュレーションの統合的活用による コンクリート内部構造診断の実現 （2）AIによる逆問題解析
英字タイトル:
A study on internal structure diagnosis of concrete structures incorporated with digital hammering inspection, AI and simulation technologies (2) Inverse problem analysis using AI
著者:
松永 嵩 小川 良太 匂坂 充行 礒部 仁博 （原燃工） 吉村 忍 山田 知典 （東京大）
発刊日:
公開日:
Concrete StructureDigital Hammering InspectionMachine LearningSimulation
A study on internal structure diagnosis of concrete structures incorporated with digital hammering inspection, AI and simulation technologies (2) Inverse problem analysis using AI
著者:
松永 嵩 小川 良太 匂坂 充行 礒部 仁博 （原燃工） 吉村 忍 山田 知典 （東京大）
発刊日:
公開日:
カテゴリ: 第17回
キーワードタグ:Concrete StructureDigital Hammering InspectionMachine LearningSimulation
概要
One of the most common inspection methods for concrete structures is hammering inspection. While this method makes it possible to quickly inspect a vast number of inspection targets, it has the problem of lacking objectivity of judgment results and recordability. In this study, we aimed to develop a technology to quantitatively diagnose the internal structure of concrete structures by digital hammering inspection. The approach of the research was to build an AI trained by CNN with a database that relates the results of digital hammering inspection to the internal structure of concrete. The constructed AI was able to estimate the size and depth of defects with sufficient accuracy.
