基于RBF网络的混凝土抗压强度的预测

为了预测混凝土的抗压强度,在分析RBF神经网络原理的基础上提出了用RBF神经网络模拟抗压强度与各影响因素间关系的方法。根据搅拌机的实际工作状况,建立了4个输入节点、1个输出节点的RBF神经网络模型,通过19组试验,验证了模型的可靠性。结果表明,实测结果与预测结果相接近,RBF神经网络模型是一种较准确的快速预测混凝土抗压强度的方法。

Prediction of Concrete Compression Strength Based on RBF Neural Network

In order to predict concrete compression strength,this paper gives out the method of using RBFneural network to express the relation of compression strength and its influencing factors after analysis of RBF neural network.Based on the working conditions of mixer,prediction model of RBF network with 4 input vectors and 1 output vectors is set up,and the model is proved by experimental results.The results show that the experimental results are closed to those of prediction and the RBF neural network is the more accurate and rapid method to predict the compression strength of concrete.