成果報告書詳細
管理番号20160000000439
タイトル*平成27年度中間年報 SIP(戦略的イノベーション創造プログラム)/インフラ維持管理・更新・マネジメント技術 維持管理ロボット・災害対応ロボットの開発 トンネル全断面点検・診断システムの研究開発
公開日2016/9/10
報告書年度2015 - 2015
委託先名東急建設株式会社 国立大学法人東京大学 学校法人湘南工科大学
プロジェクト番号P14011
部署名ロボット・AI部
和文要約
英文要約Title:Variable Guide Frame Vehicle for Inspection of Tunnel Lining Concrete
(FY2014-FY2016)FY2015 Annual Report

The current, is progressing aging was built bridges and tunnels more than 40 years ago. Inspection of the tunnel requires traffic regulation. Maintenance engineers carried out as close as possible to the inspection point by mobile elevating work platform. But, it take the time to visual inspection and hammering test for wide inspection are. And there are variations in the visual and hammering inspection by human. We have proposed a "variable guide frame vehicle for inspection of tunnel lining concrete" for new technology of infrastructure maintenance. This paper is report about the element technology of variable guide frame vehicle for inspection, each inspection system development situation.

(1) Development of Auto Crack Detection Method by Using Vision Sensors Measurement
Visible image and 3-dimensional shape depth data (range image) that obtained by light-section methods is integrated, and devised methods that distinguish crack and dirt automatic. Verification system that implemented function that extracts crack is experimented on site, and was compared to inspection results by human close visual inspection. When system and people are made comparisons, width accuracy of crack was average 108%, length accuracy of crack was 95%.
(2) Development of Auto Distinction Method of the Abnormal State Concrete
Performing diagnostic system prototype fabrication for abnormal state concrete, and performed detection experiments in concrete specimens that reproduced the diagonal cracking and flaking.
Hammering test system was conducted in 2 stage of "screening" and "details inspection", and were studied two types method of supervised multiple learning based crack depth estimation method and unsupervised learning based detection methods.
(3) Development of Variable Guide Frame Fitting for Various Tunnel Geometries
We devised a reverse analysis method of the shape and condition of the contact surface boundary of the obstacle and the VGF(Variable geometric frame). It was confirmed that in the simulated tunnel experiment can be changed shape that is possible to avoid the obstacle.
(4) Development of Inspection System that Integrates the Elements Technology
The running test that combines prototype of the traveling unit protection frame was conducted. We measured the dynamically strain of the protection frame in order to verify the structural analysis results.
(5) Development of Inspection System for the Physical Properties of Concrete
Acquisition of data for the concrete strength in piercing experiments using 2 types drills and examined the estimation method.
(6) Development of the Expert System Which Suggests the Methods for Repair of Tunnel
The questionnaire for the selection of repair technology was carried out in concrete diagnosis and maintenance engineer of 11 companies 14 people.
ダウンロード成果報告書データベース(ユーザ登録必須)から、ダウンロードしてください。

▲トップに戻る