Optimized strength modelling of foamed concrete using principal component analysis featurized regressors. (February 2023)

• Record Type: Journal Article
• Authors: Pan, Pengfei; Yang, Wenwei; Zhang, Yi; Li, Pei-Pei
• Is Part Of: Structures. Volume 48(2023)
• Page Start: 1730
• Page End: 1745
• Publication Date: 2023-02
• Subjects: Foamed concrete -- Strength modelling -- PCA -- Hyperparameter optimization -- Machine learning; Structural engineering -- Periodicals; 624.1
• DOI: 10.1016/j.istruc.2023.01.068
• Languages: English
• ISSNs: 2352-0124

Abstract

Abstract: Foamed concrete is a versatile material that can be used in different construction applications and with proper mix designing, it can also be used as a structural member. The production of sustainable lightweight foamed concrete (LWFC) requires a proper mix design relation to achieve the desired physical and mechanical properties. Numerous studies have proposed empirical formula to predict the compressive strength of foamed concrete. However, the prediction relies on a number of parameters, whose contribution to the overall material needs to be optimized. A novel parametric feature extraction using principal component analysis (PCA) is proposed for data mining. Furthermore, the accuracy of the prediction formula is limited by the amount of data with variable parameters. Due to the large differences between the calculated and actual compressive strength results, this study aims to build up a PCA-feature optimized machine learning model to enhance the design-oriented strength modelling of foamed concrete.