普通混凝土强度预测的BP神经网络模型

Prediction model of common concrete strength by BP artificial neural network

在分析普通混凝土强度影响因素基础上 ,选取混凝土配料中 7个因素作为输入值 ,混凝土2 8d强度作为输出值建立了混凝土强度预测的 BP网络模型。讨论了模型的学习样本、网络参数对预测精度的影响 ,选出最佳网络参数配置。实例证明模型预测精度高。

The influencing factors on common concrete strength were analyzed to develop the prediction model for its strength. This model is based on BP artificial neural network with 7 influencing factors as input parameters and 28 d strength as output one. The affection of learning samples and network parameters on prediction accuracy was discussed, the best network parameters were selected. A practice example shows that the prediction accuracy is high.