Scholarlink

<p>Accurately predicting bridge deck condition ratings is vital for ensuring the structural integrity and longevity of transportation infrastructure. Current evaluation methods often rely on manual inspections, which can be inconsistent, time-consuming, and prone to human error. These traditional approaches fail to provide timely and data-driven insights, which are crucial for maintenance and decision-making. This study addresses these limitations using an artificial neural network (ANN) model to predict bridge deck conditions based on nine key parameters, including bridge age, design load, and average daily traffic. The ANN model has been trained using regression techniques, with performance measured by RMSE, R<sup>2</sup>, and MAPE, achieving a strong predictive capability. Key findings highlight that bridge age significantly influences the condition rating (R<sup>2</sup> = 0.88). This model offers a more reliable, automated alternative to manual inspections, improving maintenance planning and resource allocation. Its adoption can lead to enhanced infrastructure safety and cost-efficient management.</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

Predictive assessment of bridge deck condition ratings using artificial neural networks: a data-driven approach for infrastructure management

  • 期刊论文

  • 发表日期：

    2025年7月22日

查看全文
  • Bittu Ghosh,
  • Somnath Karmakar

摘要

Accurately predicting bridge deck condition ratings is vital for ensuring the structural integrity and longevity of transportation infrastructure. Current evaluation methods often rely on manual inspections, which can be inconsistent, time-consuming, and prone to human error. These traditional approaches fail to provide timely and data-driven insights, which are crucial for maintenance and decision-making. This study addresses these limitations using an artificial neural network (ANN) model to predict bridge deck conditions based on nine key parameters, including bridge age, design load, and average daily traffic. The ANN model has been trained using regression techniques, with performance measured by RMSE, R2, and MAPE, achieving a strong predictive capability. Key findings highlight that bridge age significantly influences the condition rating (R2 = 0.88). This model offers a more reliable, automated alternative to manual inspections, improving maintenance planning and resource allocation. Its adoption can lead to enhanced infrastructure safety and cost-efficient management.