Study on Shrinkage of Concrete Based on BP Neural Network
In order to solve the problem of concrete cracking in wellhead groove with large longitudinal length, shrinkage tests of concrete containing fly ash and slag in natural environment were conducted. According to the experimental study, the shrinkage model of concrete based on BP neural network is established and the results from model agree well with the measured results. Then the effect analysis of single factor on shrinkage using the neural network model is carried out. Simultaneously, the optimal mix proportion of concrete is successfully predicted by orthogonal experiment of numerical simulation method, so that it provides reference to the control of concrete shrinkage for the engineering structure.
T. C. Wang et al., "Study on Shrinkage of Concrete Based on BP Neural Network", Advanced Materials Research, Vols. 163-167, pp. 3249-3257, 2011