Prediction of Corrosion Rate of Q235 Steel under the Marine Environment

2nd International Symposium on Resource Exploration and Environmental Science

生态环境科学

Ma, Liangtao^1 ; Dong, Haifang^1

Wuhan Second Ship Design and Research Institute, China^1

关键词: Average prediction error;    BP neural networks;    Cross-validation methods;    Environmental temperature;    Generalized regression neural networks;    Marine environment;    Metal materials;    Prediction accuracy;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/170/3/032130/pdf
DOI  :  10.1088/1755-1315/170/3/032130

学科分类：环境科学（综合）

来源: IOP

【 摘 要 】

A generalized regression neural network is used to predict the corrosion rate of metals in marine environment. The environmental temperature, oxygen content, pH value, salinity and potential are taken as input, and the corrosion rate is taken as output. A3 steel was selected to test the 25 sets of data, 18 groups of data were selected for training, and 7 sets of data were used as the verification. The results show that the generalized regression neural network prediction, select the default S value, the average prediction error is 5.72%, higher than the BP neural network was used to predict the 6.56%, using cross validation method to select the optimal S value, the optimal value of S under the average prediction error of the forecast is 2.38%. It shows that the prediction of corrosion rate of metal materials in marine environment by generalized regression neural network is feasible in technology, and has high prediction accuracy and application value.