Scholarlink

<p>The water intake of a water diversion project is generally located on the bank slope above the deep-water area of a large reservoir. The safety status of the cofferdam around the water intake during its service life is an important link to ensure the safe progress of the water diversion project. Therefore, it is necessary to carry out research on the evaluation method of its safety status. The present takes the Longhekou Reservoir Water Diversion Project as the background, and adopts a method combining Digital Image Correlation (DIC) technology, gray gradient evaluation, Swedish slice method, online monitoring platform and neural network model to design a non-contact safety status evaluation method for cofferdam structures, and carries out a systematic design on its structural monitoring and prediction methods. First, an online monitoring platform for the safety status of the cofferdam was designed based on the Swedish slice method. During the monitoring period, the platform generated more than 600,000 sets of data. When using this program to calculate the cofferdam’s safety, it only takes 20&#xa0;s to complete the safety calculation and output physical parameters such as slope safety factor, sliding surface radius, center coordinates, anti-sliding moment, sliding moment, and the angle between the radius and the horizontal plane, enabling timely early warning. Second, a safety prediction method for earth-rock cofferdams affected by physical factors was designed based on machine learning and neural networks, which can relatively predict the changes in the cofferdam’s surface displacement in the future.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A safety state assessment method for cofferdam hydraulic structures based on digital images and artificial neural network

  • 期刊论文

  • 发表日期：

    2026年2月11日

查看全文
  • Dong Chen,
  • Shan Jiang,
  • Hao Xuan,
  • Ling-wei Kong,
  • Xue-hu Zhen

摘要

The water intake of a water diversion project is generally located on the bank slope above the deep-water area of a large reservoir. The safety status of the cofferdam around the water intake during its service life is an important link to ensure the safe progress of the water diversion project. Therefore, it is necessary to carry out research on the evaluation method of its safety status. The present takes the Longhekou Reservoir Water Diversion Project as the background, and adopts a method combining Digital Image Correlation (DIC) technology, gray gradient evaluation, Swedish slice method, online monitoring platform and neural network model to design a non-contact safety status evaluation method for cofferdam structures, and carries out a systematic design on its structural monitoring and prediction methods. First, an online monitoring platform for the safety status of the cofferdam was designed based on the Swedish slice method. During the monitoring period, the platform generated more than 600,000 sets of data. When using this program to calculate the cofferdam’s safety, it only takes 20 s to complete the safety calculation and output physical parameters such as slope safety factor, sliding surface radius, center coordinates, anti-sliding moment, sliding moment, and the angle between the radius and the horizontal plane, enabling timely early warning. Second, a safety prediction method for earth-rock cofferdams affected by physical factors was designed based on machine learning and neural networks, which can relatively predict the changes in the cofferdam’s surface displacement in the future.