期刊
ADVANCES IN ENGINEERING SOFTWARE
卷 147, 期 -, 页码 -出版社
ELSEVIER SCI LTD
DOI: 10.1016/j.advengsoft.2020.102832
关键词
Machine learning; Artificial intelligence; Support vector machine; Gaussian process regression; Concrete breakout; Anchor; Fasteners; Shear Strength
Proper functioning and safety of anchor elements are decisive for the overall performance of a structural system. A possible failure mode for anchor loaded in shear is the concrete breakout failure. Concrete related failure mode poses a significant safety issue, since they may develop abruptly/brittle, without preceding signs of damage. Consequently, accurate prediction of the concrete breakout strength of anchors in shear is crucial. This study proposes two machine learning models - a Gaussian process regression (GPR) and a support vector machine (SVM) model - for predicting the concrete breakout capacity of single anchors in shear. To this end, experimental strength of 366 tests on single anchors with concrete edge breakout failures were collected from literature to establish the experimental database to train and test the models. 70% of the data were used for the model training, and the rest were used for the model testing. Shear influence factors such as the concrete strength, the anchor diameter, the embedment depth (technically the influence length), and the concrete edge distance were taken as the model input variables. The generated GPR and SVM prediction models yielded a determination coefficient R-2= 0.99 for both the training and testing data sets. Predictions from the developed models were compared to that of the other existing models (Eurocode 2, ACI 318 and Grosser) to validate their performance. The developed GPR and SVM models provided a better prediction of the experimentally observed shear strength, compared to the existing models. The predictions obtained from the GPR model are the most accurate, yielding a value of 5.6 mean absolute error when tested.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据