Scholarlink

<p>Concrete is a widely used construction material owing to its high compressive strength. However, its durability is often compromised by the development of cracks caused by tensile stress within the structures. These cracks can occur during the drying process, leading to water infiltration and corrosion of the concrete reinforcement, which subsequently require repair. Consequently, innovative technologies, such as self-repairing concrete and crack control, have become crucial for reducing the costs associated with structural repairs. In this context, this study investigated novel crack-control technologies in concrete structures using a machine-learning model that can accurately predict the performance of a specific fiber in fiber-reinforced concrete using a comprehensive dataset. The dataset was compiled from 18 studies and further augmented using synthetic data-generation techniques. It encompasses 13 different fiber types and 1953 fiber-reinforced concrete formulations. The computational model was implemented in Python, and multiple linear regression (MLR), support vector regression (SVR), random forest, and gradient boosting techniques were employed to develop the prediction model. The results showed that the Random Forest (R<sup>2</sup> = 0.887 and RMSE = 0.110), gradient boosting (R<sup>2</sup> = 0.868 and RMSE = 0.368), and SVR models (R<sup>2</sup> = 0.856 and RMSE = 0.376) outperformed their MLR counterparts (R<sup>2</sup> = 0.587 and RMSE = 0.637). Moreover, the Random Forest method showed a lower RMSE, making it more suitable for accurately predicting the performance of fiber-reinforced concrete.</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

Using synthetic data to develop machine learning models to predict the performance of fiber-reinforced concrete

  • 期刊论文

  • 发表日期：

    2025年3月11日

查看全文
  • Victor Hugo Peres Silva,
  • Carolina Luiza Emereciana Pessoa,
  • Derica dos Santos Sousa,
  • Ricardo Stefani

摘要

Concrete is a widely used construction material owing to its high compressive strength. However, its durability is often compromised by the development of cracks caused by tensile stress within the structures. These cracks can occur during the drying process, leading to water infiltration and corrosion of the concrete reinforcement, which subsequently require repair. Consequently, innovative technologies, such as self-repairing concrete and crack control, have become crucial for reducing the costs associated with structural repairs. In this context, this study investigated novel crack-control technologies in concrete structures using a machine-learning model that can accurately predict the performance of a specific fiber in fiber-reinforced concrete using a comprehensive dataset. The dataset was compiled from 18 studies and further augmented using synthetic data-generation techniques. It encompasses 13 different fiber types and 1953 fiber-reinforced concrete formulations. The computational model was implemented in Python, and multiple linear regression (MLR), support vector regression (SVR), random forest, and gradient boosting techniques were employed to develop the prediction model. The results showed that the Random Forest (R2 = 0.887 and RMSE = 0.110), gradient boosting (R2 = 0.868 and RMSE = 0.368), and SVR models (R2 = 0.856 and RMSE = 0.376) outperformed their MLR counterparts (R2 = 0.587 and RMSE = 0.637). Moreover, the Random Forest method showed a lower RMSE, making it more suitable for accurately predicting the performance of fiber-reinforced concrete.