人工神经网络在掺活化煤矸石粉混凝土强度预测中的应用  

Application of artificial neural network in the prediction of the compressive strength of concrete with activated coal gangue powder

作　　者：周双喜[1] 赵碧华[1] 

机构地区：[1]华东交通大学土木工程学院,江西南昌330013

出　　处：《混凝土》2007年第4期26-28,共3页Concrete

摘　　要：人工神经网络技术综合考虑了掺活化煤矸石混凝土强度的各种影响因素,可用于预测混凝土强度。选取了掺活化煤矸石粉混凝土配料中7个主要因素作为输入值,混凝土28d强度作为输出值,建立起混凝土强度预测BP网络模型,进而对掺活化煤矸石配合比强度试验数据进行分析预测,效果良好。结果表明该方法用于掺矿物掺合料混凝土强度预测方面是可行的。Artificial neural network predict the compression strength of concrete with activated coal gangue powder,general considering rarious influencing factors. Artificial neural network can deal with non-line problem and predict strength. This prediction model is based on BP artificial network with 7 influencing factors as input parameters and 28 days strength as output one. The analysis and test on data about mixed ration and strength of concrete have been performed, too. The results show it is very useful to predict the strength of concrete with mineral additives in these methods.

关 键 词：人工神经网络 活化煤矸石粉 混凝土强度 

分 类 号：TU528.041[建筑科学—建筑技术科学]