|
|
PREDICTING CORROSION RATE OF MILD STEEL IN SOIL BASED ON ARTIFICIAL NEURAL NETWORK |
GUO Zhihu; XING Zhengliang; JIN Minghui; MENG Xialan (Huazhong University of Science and Technology; Wuhan 430074) |
|
|
Abstract Artificial neural network, possessing learning and non-linear character, was applied to study corrosion of mild steel in soil. A neural network typically consists of many simple neurons like processing elements called "cell" or "nodes" that interact with other cells using numerical weighted connection. In this study,a neural network with 6-10-1 structure, namely 6 input nodes, 10 hidden layer nodes, 1 output node, was used. The learning algorithm was BP (Back-Propagation) algorithm. Corrosion tests of mild steel in soil were carried out with orthogonal test method, in which five corrosion factors, namely pH value,Cl-, H2O,SO_4~2- and Fe2+ content, were considered and test data of sixteen groups were obtained. These data were used as sample set to train neural network. The inputs of neural network were the five corrosion factors and test duration, the output of neural network was corrosion rate of steel in soil.The research results showed that soil corrosion rates of mild steel in 24 months could be predicted by the trained artificial neural network, and were basically in agreement with experimental data.
|
Received: 25 August 1996
|
Cite this article:
GUO Zhihu; XING Zhengliang; JIN Minghui; MENG Xialan (Huazhong University of Science and Technology; Wuhan 430074). PREDICTING CORROSION RATE OF MILD STEEL IN SOIL BASED ON ARTIFICIAL NEURAL NETWORK. J Chin Soc Corr Pro, 1996, 16(4): 307-310.
URL:
https://www.jcscp.org/EN/ OR https://www.jcscp.org/EN/Y1996/V16/I4/307
|
1郭稚孤,金名惠,桂修文,孟厦兰,邢政良,罗逸.腐蚀科学与防护技术,1995;7(3):2582尹红风,戴汝为.模式识别与人工智能,1990,3(1):13王强编著.地下金属管道的腐蚀与阴极保护,西宁:青海人民出版社,1984.404KingRA,MillerJDA,WakerleyDS.Br.Corros.J,1973,8(3):895郭稚弧,唐和清,张君华,张维宇,佐佐本和夫.广岛大学工学部研究报告,1992;40(1):516周建华.介质条件对碳钢在上壤中腐蚀速度的影响,华中理工大学化学系学士论文,1993 |
No Suggested Reading articles found! |
|
|
Viewed |
|
|
|
Full text
|
|
|
|
|
Abstract
|
|
|
|
|
Cited |
|
|
|
|
|
Shared |
|
|
|
|
|
Discussed |
|
|
|
|