Title: Estimating the compressive strength of concrete using multiple linear regression and adaptive neuro-fuzzy inference system
Authors: Faezehossadat Khademi; Sayed Mohammadmehdi Jamal
Addresses: Illinois Institute of Technology, Chicago, IL, USA ' University of Hormozgan, Bandar Abbas, Iran
Abstract: Evaluating the concrete quality is a significant factor in the concrete industry. Concrete compressive strength, identified as one of the most important mechanical properties of concrete, is recognised as the most essential parameter for the quality assurance of concrete. In this paper, in order to evaluate the 28-day compressive strength of concrete, the two most challenging models of multiple linear regression (MLR) and adaptive neuro-fuzzy inference system (ANFIS) are developed in MATLAB environment for 160 different concrete specimens and the results are compared with each other. The results indicate that ANFIS model could perfectly predict the compressive strength of concrete; however, multiple linear regression model was not as effective as ANFIS in predicting purposes. The superiority of ANFIS to MLR might be because of the nonlinear relationships between the concrete characteristics which ANFIS is more capable in their modelling purposes.
Keywords: multiple linear regression; MLR; adaptive neuro-fuzzy inference system; ANFIS; concrete quality; compressive strength; neural networks; fuzzy logic; quality assurance; modelling; simulation.
International Journal of Structural Engineering, 2017 Vol.8 No.1, pp.20 - 31
Received: 12 Feb 2016
Accepted: 15 Aug 2016
Published online: 19 Jan 2017 *