نوع مقاله : پژوهشی
نویسندگان
دانشکده فنی و مهندسی، دانشکده ی عمران - دانشگاه رازی
چکیده
کلیدواژهها
عنوان مقاله [English]
نویسندگان [English]
In the present paper, artificial neural networks (ANN) and regression analysis for predicting compressive strength of cubes ofconcrete containing silica fume (SF), fly ash, and Copper slag are developed at the age of 7 and 28 days. For building these models, training and testing using the available experimental results for 66 specimens produced with 6 different mixture proportions are used. The data used in the multi-layer feed forward neural networks models and linear regression model are designed in the format of seven input parameters covering the age of specimen, cement, fine aggregate, coarse aggregate, fly ash, silica fume, and copper slag. According to these input parameters, in the multi-layer feed forward neural networks, models are used to predict the compressive strength and durability values of concrete. It was shown that neural networks have high potential for predicting the compressive strength and durability values of the concretes containing silica fume (SF), fly ash and copper slag. Results show that the values obtained from the training and testing in ANN-I (LM Algorithm) model are very closer to the experimental results. The results show that ANN has strong potential as a feasible tool for estimating the ingredients of concrete to meet the design requirements. Also, multiple regression (MR) is a statistical technique that allows us to predict someone's score on one variable on the basis of their scores on several other variables. MR is employed to learn more about the relationship between several independent or predictor variables and a dependent or criterion variable. Therefore, MR analysis was carried out using a MATLAB 2013 package to correlate determined fc value to the seven concrete parameters. The data used while developing the ANN model (i.e., 66 data sets) were used in the development of the MR model. However, the obtained indices make it clear that the ANN model is more capable with a higher prediction performance compared to the MR model.
کلیدواژهها [English]
)