Methodology of Neural Identification of Strength of Concrete

In this article, a new technique of the neural identification of the compression strength of concrete on the basis of nondestructive tests is described. The neural training and testing processes were carried out on a data set created on the basis of nondestructive tests. The data set covers a concrete compression strength range from 24 to 105 MPa after 28 days of curing. The methodology of the neural identification of the compression strength of concrete is presented. The results presented herein demonstrate that the identification of compressive strength of concrete by neural networks on the basis of such nondestructive methods as ultrasonic, sclerometric, and pull-out is a viable technique.