Dam & Safety ›› 2022, Vol. 0 ›› Issue (4): 23-.

Previous Articles     Next Articles

Research on intelligent pattern classification of dam safety monitoring data

SUN Futing and SHEN Haiyao   

  1. Large Dam Safety Supervision Center, National Energy Administration
  • Received:2022-03-17 Online:2022-08-08 Published:2022-10-21

大坝安全监测数据模式智能识别技术研究

孙辅庭,沈海尧   

  1. 国家能源局大坝安全监察中心,浙江  杭州,311122
  • 作者简介:孙辅庭(1987— ),男,浙江杭州人,高级工程师,主要从事大坝安全监控、水工及岩体结构数值仿真、人工智能技术应用工作。

Abstract: Intelligent pattern classification of dam safety monitoring data is researched. Combining with the characteristics of engineers' thinking and deep learning, the technical scheme of data pattern classification is proposed. On this basis, the pattern classification model of dam safety monitoring data is established based on convolutional neural network. The case study shows that the proposed algorithm performs well in pattern classification of monitoring data. Compared with previous methods, this method has the advantage of using one model to accurately and quickly process massive and multi-type monitoring data without artificial definition of mathematical model. This research is the extension of application of the new generation of artificial intelligence technology in the field of dam safety, which is also the key technology to realize intelligent management of dam safety.

Key words: dam safety, monitoring, pattern recognition, CNN, artificial intelligence

摘要: 对大坝安全监测数据规律性智能识别技术开展研究。结合工程师思维和深度学习的特点,提出数据规律性识别技术方案,在此基础上,基于卷积神经网络建立大坝安全监测数据规律性识别模型。算例研究表明,建立的智能识别算法在监测数据规律性识别问题上具有较好表现,与以往方法相比,无需人为定义数学模型,具备采用一个模型准确快速处理海量多类型监测数据的优势。研究成果是新一代人工智能技术在大坝安全领域应用的拓展,是实现大坝安全智能管理的关键技术点。

关键词: 大坝安全, 监测, 模式识别, 卷积神经网络, 人工智能

CLC Number: