High-performance Concrete Strength Prediction Model Based on the Radial Basis Function Neural Network of Human Cerebral Cortex

摘要

Artificial neural network (ANN) is a kind of bionic technology which stimulates the information processing mechanism of neural transmission in human cerebral cortex, using this technology we can process complex non-linear system with multiple influencing factors, such as the concrete system. Based on this, this paper proposes a prediction model of high strength concrete based on radial basis artificial neural network. Through empirical analysis, it is found that the ANN prediction model has fast training speed, good prediction effect, and high prediction accuracy. The prediction accuracy of RBF neural network prediction model is better than that of GRNN neural network model. The average absolute error is 1.04MPa, the average absolute percentage error is 2.09%, and the mean square error is 1.75MPa, which can meet the requirements in the construction projects.