首页|Studies from Charles Darwin University Add New Findings in the Area of Machine Learning (Tree-based Machine Learning Approach To Modelling Tensile Strength Retention of Fibre Reinforced Polymer Composites Exposed To Elevated Temperatures)
Studies from Charles Darwin University Add New Findings in the Area of Machine Learning (Tree-based Machine Learning Approach To Modelling Tensile Strength Retention of Fibre Reinforced Polymer Composites Exposed To Elevated Temperatures)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
Elsevier
Research findings on Machine Learning are discussed in a new report. According to news originating from Darwin, Australia, by NewsRx correspondents, research stated, "Fibre Reinforced Polymer (FRP) composites are susceptible to degradation at elevated temperatures. Accurate modelling of the tensile performance of FRP composites under high-temperature exposure is crucial for their structural integrity." Our news journalists obtained a quote from the research from Charles Darwin University, "In this study, tree-based models, namely, decision tree, M5P, and random forest methods, are utilised to model the impact of elevated temperatures on the tensile strength of composite materials. A database of 787 experimental results is established and processed to train and test the regression tree models. The exposure temperature, resin glass transition temperature, sample thickness/diameter, exposure duration, ambient cooling, fibre-toresin ratio, fibre orientation, resin type, fibre type, and manufacturing process were considered as the main parameters affecting the tensile strength retention (TSR) of FRP composites after exposure to elevated temperatures. To improve the prediction performance of machine learning, Bayesian optimisation and 10- fold cross validation (CV) technique were used to train regression tree methods. The results demonstrated the accuracy of the developed models in predicting the TSR of the composites under elevated temperatures. Feature contribution analysis showed that the exposure temperature exerts the most significant impact on the TSR, with the glass transition temperature coming next in importance. These were followed by sample thickness, exposure duration, ambient cooling, fibre-to-resin ratio, and fibre orientation, respectively. Resin type, fibre type, and the manufacturing process had the least contributions to the observed variations in TSR."
DarwinAustraliaAustralia and New ZealandCyborgsEmerging TechnologiesMachine LearningCharles Darwin University
NETL
NSTL
Elsevier
