摘要
通过模糊聚类确定了影响碳钢腐蚀速率的主要土壤因素,对14个腐蚀站点土壤的理化性质以及碳钢的年腐蚀数据进行分析,构建了碳钢年土壤腐蚀预测模型,利用该模型在BP人工神经网络中进行学习、训练、模拟,并将预测结果与现场碳钢埋片腐蚀实验结果对比。结果表明:含水量、pH值、Cl-含量、SO42-、电导率、可溶盐总量6种土壤环境参数为影响14个腐蚀站点土壤中碳钢腐蚀的主要因素;运用BP人工神经网络可以建立起稳定性好的土壤腐蚀预测模型,较好地预测了我国典型地区碳钢在土壤中的腐蚀速率。
The main environmental factors of soil corrosion for carbon steels in regional soils of 14 test cities in China were determined using fuzzy clustering. By analyzing the physical and chemical properties of soil and annual corrosion data of carbon steel, the soil corrosion prediction model for carbon steel was built. The reasonableness of the corrosion model was verified by using the BP artificial neural network to learn, train, simulate, and compare with the corrosion test results of the carbon steel samples buried in 14 cities regional soil. The results show that water content, pH, C1- content, SO42- content, soil conductivity and total dissolved salts are the six main factors on soil corrosion of carbon steel in the 14 cities regional soil, and a stable forecasting model can be built based on the BP artificial neural network, which well predicted the corrosion rates of carbon steel in regional soils of 14 test sities.
出处
《腐蚀科学与防护技术》
CAS
CSCD
北大核心
2013年第5期372-376,共5页
Corrosion Science and Protection Technology
基金
河北省自然科学基金项目(E2013402004)资助
关键词
碳钢
土壤腐蚀
BP人工神经网络
腐蚀速率
carbon steel, soil corrosion, BP artificial neural network, corrosion rate
