Scholarlink

<p>This study focuses on the critical aspect of skid resistance in asphalt pavements, an essential factor for ensuring road safety. Several factors, including pavement texture, aggregate properties, and environmental conditions, influence skid resistance. Utilizing data from the Long-Term Pavement Performance (LTPP) database, this research develops an Artificial Neural Network (ANN) model to predict skid resistance based on a comprehensive array of inputs such as traffic volume, pavement age, asphalt content, and environmental factors. A key outcome of this study is the derivation of an explicit algebraic equation that fully represents the trained ANN model, allowing practitioners to compute Friction Number (FN) without running the whole model. The model was trained with data involving 135 GPS pavement sections across various climate zones in the United States, achieving a robust Coefficient of Determination (R²) of 0.87, a Root Mean Square Error (RMSE) of 2.52, a Mean Absolute Error (MAE) of 2.04, and a Mean Absolute Percentage Error (MAPE) of 4.4%. The findings underscore the significant impact of traffic loads, pavement age, and environmental conditions on skid resistance. The study’s methodology and findings contribute to the broader understanding of pavement management and safety strategies, providing a valuable tool for predicting and maintaining optimal skid resistance in asphalt pavements. Recommendations for future research include expanding the dataset and refining the model with additional influential factors to enhance prediction accuracy and reliability.</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

Enhancing road safety: developing a neural network-based model for predicting skid resistance in asphalt pavements using LTPP data

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年12月23日

查看全文
  • Tanvir Ahmed,
  • Mayzan Isied,
  • Mena Souliman

摘要

This study focuses on the critical aspect of skid resistance in asphalt pavements, an essential factor for ensuring road safety. Several factors, including pavement texture, aggregate properties, and environmental conditions, influence skid resistance. Utilizing data from the Long-Term Pavement Performance (LTPP) database, this research develops an Artificial Neural Network (ANN) model to predict skid resistance based on a comprehensive array of inputs such as traffic volume, pavement age, asphalt content, and environmental factors. A key outcome of this study is the derivation of an explicit algebraic equation that fully represents the trained ANN model, allowing practitioners to compute Friction Number (FN) without running the whole model. The model was trained with data involving 135 GPS pavement sections across various climate zones in the United States, achieving a robust Coefficient of Determination (R²) of 0.87, a Root Mean Square Error (RMSE) of 2.52, a Mean Absolute Error (MAE) of 2.04, and a Mean Absolute Percentage Error (MAPE) of 4.4%. The findings underscore the significant impact of traffic loads, pavement age, and environmental conditions on skid resistance. The study’s methodology and findings contribute to the broader understanding of pavement management and safety strategies, providing a valuable tool for predicting and maintaining optimal skid resistance in asphalt pavements. Recommendations for future research include expanding the dataset and refining the model with additional influential factors to enhance prediction accuracy and reliability.