BP神经网络在路面材料抗压强度预测中的应用

应用MATLAB提供的神经网络工具箱作为BP神经网络训练和仿真的平台，并进行语言编程，通过采用不同隐函数节点数进行对比试验，采用精度与误差都合适的节点数进行训练与预测，观察预测的精度，并分析神经网络对抗压强度结果预测的可应用性，从而得出一些有益的结论。

Application of BP neural network in the prediction of pavement materials compressive strength

Applied neural network toolbox provided by MATLAB as BP neural network training and simulation platform, and made programming language. By adopting different implicit function node number to make comparative tests, adopted precision and error were all suitable nodes to make training and prediction. Observed the prediction accuracy, and analysed the application of neural network to compressive strength prediction results, so as to draw some useful conclusions.