Scholarlink

<p>The structural integrity of concrete infrastructures is a paramount concern in civil engineering. Traditional methods for identifying different damages/defects are often hindered by their invasive nature, subjectivity, and labor intensity. This study introduces a novel approach to real-time detection and localization of such damages/defects which are equivalent to initial cracks in concrete structures by employing a Multilayer Perceptron (MLP) model to analyze Acoustic Emission (AE) data. Focused on the development of a non-invasive, automated monitoring system, the research aims to surpass the limitations of traditional inspection techniques with a machine-learning framework that enhances detection accuracy and efficiency. Utilizing AE signals from simulated cracks in a reinforced concrete slab, the MLP model was trained to identify and localize structural damage. The findings demonstrate the model’s superior performance in detecting incipient cracks, offering a significant advancement in the field of structural health monitoring. This innovation promises to improve maintenance strategies and risk management for infrastructure safety, marking a pivotal step toward intelligent and resilient civil engineering practices.</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

Crack localization in RCC slabs using acoustic emission data and multilayer perceptron

  • 期刊论文

  • 发表日期：

    2025年11月5日

查看全文
  • Soumyadip Das,
  • Aloke Kumar Datta,
  • Pijush Topdar,
  • Gour Sundar Mitra Thakur,
  • Sanjay Sengupta

摘要

The structural integrity of concrete infrastructures is a paramount concern in civil engineering. Traditional methods for identifying different damages/defects are often hindered by their invasive nature, subjectivity, and labor intensity. This study introduces a novel approach to real-time detection and localization of such damages/defects which are equivalent to initial cracks in concrete structures by employing a Multilayer Perceptron (MLP) model to analyze Acoustic Emission (AE) data. Focused on the development of a non-invasive, automated monitoring system, the research aims to surpass the limitations of traditional inspection techniques with a machine-learning framework that enhances detection accuracy and efficiency. Utilizing AE signals from simulated cracks in a reinforced concrete slab, the MLP model was trained to identify and localize structural damage. The findings demonstrate the model’s superior performance in detecting incipient cracks, offering a significant advancement in the field of structural health monitoring. This innovation promises to improve maintenance strategies and risk management for infrastructure safety, marking a pivotal step toward intelligent and resilient civil engineering practices.