Concrete in the low temperature area will freeze and thaw with the change of seasons,and influence its durability.The proportion of raw materials has an important effect on the frost resistance of concrete.By optimizing the ratio of raw materials,the freezing resistance of concrete can be improved,and its service life can be extended.The influence of raw material matching ratio on concrete is calculated by using grey correlation degree theory,and the mechanical properties of concrete are predicted by BP neural network.Variables,such as cement,limestone powder,fly ash,slag,fine aggregate,coarse aggregate,water,water reducing agent and cycle times are taken as input factors,and compressive strength is taken as output variables.The results show that there is high correlation between slag,fine aggregate,coarse aggregate,water,cycle times and freezing resistance,while the other raw material content has relatively little effect on freezing resistance.Adjusting the proportion of raw materials can improve the mechanical properties and frost resistance of concrete.BP neural network can effectively predict the compressive strength.The prediction error is small,and the correlation coefficient is larger than 0.96.This result can effectively reduce the workload of concrete experimental research,and reduce the engineering cost.
关键词
混凝土/冻融循环/灰色关联度理论/BP神经网络
Key words
Concrete/Cycle of freezing and thawing/Grey correlation analysis/BP neural network
维普
万方数据