Compressive strength prediction of high-strength oil palm shell lightweight aggregate concrete using machine learning methods

Abstract

Promoting the use of agricultural wastes/byproducts in concrete production can significantly reduce environmental effects and contribute to sustainable development. Several experimental investigations on such concrete's compressive strength ([Formula: see text]) and behavior have been done. The results of 229 concrete samples made by oil palm shell ([Formula: see text]) as a lightweight aggregate ([Formula: see text]) were used to develop models for predicting the [Formula: see text] of the high-strength lightweight aggregate concrete ([Formula: see text]). To this end, gene expression programming ([Formula: see text]), adaptive neuro-fuzzy inference system ([Formula: see text]), artificial neural network ([Formula: see text]), and multiple linear regression ([Formula: see text]) are employed as machine learning ([Formula: see text]) and regression methods. The water-to-binder ([Formula: see text]) ratio, ordinary Portland cement ([Formula: see text]), fly ash ([Formula: see text]), silica fume ([Formula: see text]), fine aggregate ([Formula: see text]), natural coarse aggregate ([Formula: see text]), [Formula: see text], superplasticizer ([Formula: see text]) contents, and specimen age are among the nine input parameters used in the developed models. The results show that all [Formula: see text]-based models efficiently predict the [Formula: see text]'s [Formula: see text], which comprised [Formula: see text] agricultural wastes. According to the results, the [Formula: see text] model outperformed the [Formula: see text] and [Formula: see text] models. Moreover, an uncertainty analysis through the Monte Carlo simulation (MCS) method was applied to the prediction results. The growing demand for sustainable development and the crucial role of eco-friendly concrete in the construction industry can pave the way for further application of the developed models due to their superior robustness and high accuracy in future codes of practice.

Keywords: Agricultural waste; High-strength concrete; Lightweight aggregate concrete; Machine learning; Strength prediction.