Application of computational intelligence and statistical approaches for auto-estimating the compressive strength of plastic concrete

Abstract

The use of clays, especially bentonite clays in composites, is one of the practical and cost-effective ways for engineers to absorb heavy metals. On the other hand, automatically obtained mechanical properties of these composites, especially plastic concrete (PC) made by adding bentonite clay to the ordinary concrete mixture, can be useful in reducing the cost and optimising the time of the cutoff wall construction in earth dam projects. For this purpose, four computational intelligence and statistical approaches, including support vector machine (SVM), multi-gene genetic programming (MGGP), group method of data handling (GMDH), and response surface methodology (RSM), were used in this article to auto-estimate the compressive strength (CS) of PC. To do this, a comprehensive dataset published in the research literature was used. Also, the models validity was assessed by both external validations of the proposed approaches and Monte-Carlo uncertainty analysis. Finally, each input variable in the models has been studied parametrically over wide ranges to provide practical results to engineers. Comparing the results with the existing literature indicates the acceptable accuracy of the developed models in this study due to optimisation in the speed of developed models and also the less complexity of the proposed ones.