Research on Analysis of Infiltration of Stormwater Volume by AI Machine Learning of Flow Rate, Water Temperature Method Data

DOI Web Site Open Access
  • Sato Katsumi
    日本大学生産工学部 土木工学科 日本下水道協会 特別会員
  • Nakane Susumu
    中日本建設コンサルタント株式会社 水工技術本部 日本下水道協会 特別会員
  • Takahashi Iwahito
    日本大学生産工学部 土木工学科 日本下水道協会 特別会員
  • Hosaka Seiji
    日本大学生産工学部 環境安全工学科 日本下水道協会 特別会員
  • Morita Hiroaki
    日本大学生産工学部 土木工学科 日本下水道協会 特別会員

Bibliographic Information

Other Title
  • 流量・水温法データのAI機械学習による雨天時浸入水量解析の研究
  • リュウリョウ ・ スイオンホウ データ ノ AI キカイ ガクシュウ ニ ヨル ウテンジ シンニュウ スイリョウ カイセキ ノ ケンキュウ

Search this article

Abstract

<p>　The authors estimated the decrease in water temperature due to precipitation from the measured water temperature using a neural network, which is one of AI machine learning. It was shown that the non-excess probability represents a decrease in water temperature due to precipitation. As a result of measuring the water temperature and flow rate at the same point and analyzing the infiltration of storm water rate using this data, it was confirmed that there is a correlation with the non-excess probability. Then, it was shown that the non-excess probability represents the infiltration of stormwater rate ratio. In this study, the flow rate measured using machine learning was used as teacher data, and the estimated flow rate and the estimated flow rate in fine weather were derived from it. Then, it was confirmed that the neural network method can be reproduced best and is effective for this analysis. We also considered the selection of explanatory variables suitable for water temperature analysis, the appropriate measurement interval of water temperature data, and the survey area. At the same time, it was shown that the amount of infiltrated of stormwater between measurement points, can be analyzed by measuring either the flow rate or the water temperature.</p>

Journal

Related Projects

See more

Keywords

Details 詳細情報について

Report a problem

Back to top