本文へジャンプ

成果報告書詳細
管理番号20170000000608
タイトル*平成28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/インフラの多種多様なセンシングデータを処理・蓄積・解析する技術の開発/インフラセンシングデータの統合的データマネジメント基盤の研究開発
公開日2018/3/8
報告書年度2016 - 2016
委託先名大学共同利用機関法人情報・システム研究機構国立情報学研究所 国立大学法人北海道大学
プロジェクト番号P14031
部署名IoT推進部
和文要約
英文要約Title: An Integrated Data Management Platform for Civil Infrastructure Sensing (FY2014-2018) FY2016 Annual Report

1.1 Sensing Database System
A prototype system for large sensor data management was developed. It is built on a distributed NoSQL database management system that enables scalable and flexible sensor data management. We evaluated the performance of the system and showed that it can retrieve any fragment of stored sensor data efficiently. We also developed a video analysis system for extracting vehicle images from surveillance video data. The system extracts vehicles by background image processing and vehicle recognition in video frames. It was confirmed that the system could extract vehicle images accurately when we checked surveillance video data during sunny daytime on Myoko Bridge.

1.2 Visual Analytics System
An exploratory visual analytics system with multiple views was developed. First, a calculus system was constructed to calculate statistical values such as mean, maximum, RMS and analytic scores such as number of abnormal vibrations. It uses long-term data from three-point accelerometer sensors as well as meteorological data such as inside/outside atmospheric temperatures of a bridge. Second, an analytic system was constructed which can detect anomalous vibrations and stress distribution changes in a bridge structure using data visualization of multiple real-time synchronous data streams. The developed system is available at the open source software repository GitHub (https://github.com/truemrwalker/wblwrld3/).

1.3 High-precision Time Synchronization Multi-Sensing Technology
A multi-sensing technology was developed that secures time synchronicity without GPS and various communications by giving time information using a chip scale atomic clock (CSAC). We implemented an autonomous time synchronization multi-sensing practical module on Myoko Bridge and the Central Expressway and made improvements. The specified performance was confirmed against the following improvements: expansion of external analog sensor input interface to 3 channels, enhanced AD converter to 24 bits, enhanced FPGA, wireless communication function realized with commercially available Raspberry Pi and time synchronization by IEEE 1588.

1.4 Data mining for bridge monitoring data via collaboration between Information Technology and Civil Engineering
The relation between atmospheric temperature and the natural frequency was extracted. We found the natural frequencies at Myoko Bridge seemed not to depend on temperature so large compared with other steel-type bridges. This characteristic of the natural frequency is good for an index characterizing the bridge condition. However, it is known that the natural frequency is insensitive to damages of a bridge, as its initial stiffness does not change unless a designed live load is put on it. Thus, we have also been searching for characteristic values having sensitivity for damages using numerical simulations.
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