Prediction of Compressive Strength of Concrete using Artificial Neural Networks

The primary composition of concrete includes cement, water and aggregates. The main objective in proportioning of these ingredients is to produce concrete of desired strength. Concrete being a complex material, the prediction of compressive strength is a cumbersome task. In this study, Artificial intelligence model is put forth to predict the strengths at various ages of concrete which will definitely save time, material and money. Artificial neural networks are gaining popularity and have proved to be a promising area of Artificial Intelligence. Artificial neural networks derive their origin from human brain. The use of this technology where a computer is used to mimic large amount of interconnections and networking that exists between nerve cells like in human nervous system has proved to be an efficient one. The proposed model has inputs namely cement, sand, water, coarse aggregates, fine aggregates and fineness modulus. Present study involves data obtained and the network is trained using a back propagation algorithm. This algorithm uses layered feed forward artificial neural networks. Further this algorithm is a supervised learning method which is a generalization of delta rule and is activated by log-sigmoidal function.