基于BP网络的再生混凝土抗压强度的预测模型

Forecast model about compressive strength of recycle aggregate concrete base on BP neutral network

利用BP神经网络法,建立了再生混凝土的抗压强度预测模型。该模型采用了3层网络结构模式,输入层采用再生混凝土的配合比数据,输出层为再生混凝土,置放7、28、56、90d的强度数据,模型的转移函数均采用单极性Sigmoid函数。由于Sigmoid函数值为[0,1],因此,对再生混凝土输入数据进行归一化处理。最后,设计了33块再生混凝土的抗压强度试验,利用试验数据对网络模型进行测试,测试的结果证实了该模型对再生混凝土的强度预测值与实际测试结果基本相符。

The forecast model about the compressive strength of recycled aggregate concrete was set up using the method of BP neutral network.The model was three-layer network structure pattern.The inputting-layer was the mix ratio of recycled aggregate concrete.The outputting-layer was the strengths data of the recycled concentrate in 7,28,56,90 d.The model＇s transfer function was Sigmoid function.Because the numerical value of Sigmoid was in [0,1].This article turned the inputting data into the data in [0,1].At last,the experimental work for the compressive strength of recycled aggregate concrete was designed and performed with 33 pieces.The model was trained and tested by experimental data.It showed that forecast result was match with test results and it could be referenced for experimentation and manufacture.