| 引用本文: | 兰志刚,侯保荣,白 刚,宋积文,陈胜利,谭 震,张 杰.海洋环境中平台钢腐蚀速率的三层BP 神经网络预测[J].海洋科学,2010,34(12):75-77. |
| |
|
| 摘要: |
| 利用三层BP 神经网络预测海洋环境因素对材料的腐蚀速率的影响。结合实测的pH 值、温度、溶解氧、盐度、生物附着等影响因素, 分析了上述环境因素对平台钢腐蚀的影响, 建立环境因素与腐蚀速率之间的映射关系, 预测了平台钢在海洋环境中的腐蚀速率。结果表明, 全浸区腐蚀速率预测误差为6.95%, 潮差带腐蚀速率预测误差为4.2%, 预测精度较高。说明利用三层BP 神经网络预测钢在海水中腐蚀速率技术可行, 具有较高的预测精度和应用价值。 |
| 关键词: 腐蚀因素 腐蚀速率预测 BP 神经网络 海洋环境腐蚀预测 |
| DOI： |
| 分类号: |
| 基金项目:海洋石油总公司综合科研项目 |
|
| Prediction of effects of marine environmental factors on steel corrosion rates with three-layer BP neural network |
|
|
| Abstract: |
| We introduced the methodology to study relationship between steel corrosion and marine environmental factors and to predict of steel corrosion rates with three-layer BP neural network. With the in situ measurements of pHs, water temperatures, dissolved oxygen, salinities and bio-fouling, the effects of marine environmental factors on steel corrosion were analyzed and sorted in a descending sequence. With a three-layer BP neural network, the corrosion rates of steel in seawater were predicted with an error of 6.95% in submerged zones and 4.2% in tidal zones. The results show that prediction with the neural network was feasible, producing good prediction accuracy and value. |
| Key words: corrosion factors prediction of corrosion rate three-layer BP neural network prediction of corrosion on marine environmental |
