基于神经网络的水泥强度预测

Predicting of cement strength based on neural network

由于水泥强度预测具有多变量、非线性、大时滞的特点 ,采用传统的线性回归分析预测法与聚类分析预测法 ,预测结果的准确性较低 ,文章利用神经网络的算法 ,建立了水泥强度的动量 -自适应神经网络预测模型。通过对有关数据进行实例计算 ,得到了令人满意的结果 ,该模型为水泥强度的快速预测提供了一种新方法 。

The conventional methods of predicting cement strength by linear regression analysis and cluster analysis show a low accuracy in prediction results.In this paper, the momentum and self adaptive ANN predicting model of cement strength is made based on the characteristics of multi variable, non linearity and long time lag in the prediction of cement strength.The calculation of data taken from the related literature is carried out,and a satisfactory result is obtained. The presented model provides a new way of predicting cement strength quickly,which is practically valuable.