基于人工神经网络模型的C25混凝土配合比的预测方法

Neural Network Based Prediction of Mixing Ratio for C25 Concrete

根据混凝土强度早期判定的特点 ,建立了能映射复杂非线性关系的多层前向神经网络 .并且利用前向神经网络进行混凝土的配合比设计预测 ,结果表明神经网络方法具有较高的预测精度 。

This paper discussed characteristics of concrete strength prediction, and on the basis of the analysis, built a multi layer feed forward neural network model to perform complex non linear mapping. Feed forward network is applied to research and design the mixing ratio for C25 concrete. The results suggest that neural network approach has merit of high forecasting accuracy. So it'll have broad prospect of application in performance forecast and optimal design of concrete.