摘要
利用灰色关联度法对不同强度等级不同腐蚀环境下混凝土抗腐蚀性进行分析。利用神经网络对混凝土抗腐蚀性试验数据进行训练,并对未参与训练的数据进行预测验证。研究结果表明:砂子、水泥、碎石及防腐涂层等4种因素与抗腐蚀性的相关性最高,其余影响因素对抗腐蚀性影响则相对较小。BP神经网络预测的抗腐蚀性参数与试验所测数据误差基本在10%以内,满足要求。
The corrosion resistance of concrete under different strength levels and different corrosion environments was analyzed by using grey relational degree method.The neural network is used to train the concrete corrosion resistance test data,and to predict and verify the data that is not involved in the training.The results show that sand,cement,gravel and anti-corrosion coating have the highest correlation with corrosion resistance,while the other factors have relatively little effect on corrosion resistance.The error between the corrosion resistance parameters predicted by BP neural network and those measured by experiment is basically less than 10%,which can basically meet the requirements.The test and prediction results can provide reference for the prediction of compressive strength of concrete under different strength grades and different corrosion environments in saline soil area.
作者
李渊
景涛
申铁军
LI Yuan;JING Tao;SHEN Tie-jun(Taiyuan Coal Industry Taiyuan Design and Research Institute Group Co.,LTD,Taiyuan,Shanxi,030024,China;Shanxi Road&Bridge Construction Group Co.,Ltd.,Taiyuan,Shanxi,030006,China)
出处
《建材技术与应用》
2024年第2期18-21,共4页
Research and Application of Building Materials
关键词
抗腐蚀
灰色关联度
BP神经网络
抗压强度
corrosion resistance
grey correlation degree
BP neural network
compressive strength
