基于KPCA-ICS-ELM模型的油气水混输管道腐蚀速率预测分析  

Analysis on Corrosion Rate Prediction of Oil-Gas-Water Mixed Transportation Pipeline Based on KPCA-ICS-ELM Model

作　　者：郭永强[1] 李希[1] 周忠强[1] 李文鹏[1] 杨斌[1] 姚瑞峰 GUO Yongqiang;LI Xi;ZHOU Zhongqiang;LI Wenpeng;YANG Bin;YAO Ruifeng(Third Gas Production Plant,Changqing Oilfield of PCCL,Wushenqi 017300,China)

机构地区：[1]中国石油天然气股份有限公司长庆油田分公司第三采气厂,内蒙古乌审旗017300

出　　处：《热加工工艺》2022年第16期54-59,63,共7页Hot Working Technology

基　　金：中石油重大科技专项资助项目(2018E-1809)。

摘　　要：由于油气水混输管道内的腐蚀性物质含量较多,因此,腐蚀速率相对较快。研究提出了一种基于KPCA-ICS-ELM模型的腐蚀速率预测模型,首先对KPCA(核主成分分析)算法、ICS(改进布谷鸟搜索)算法、ELM(极限学习机)算法分别进行介绍,提出KPCA-ICS-ELM模型的组合方法以及模型评估方法,在此基础上,通过实验的方式获取油气水混输管道的腐蚀速率数据,进而构建指标体系,通过实例验证的方式证明本次所提模型的可行性,通过与其它模型进行对比,证明此模型的先进性。结果表明:通过使用KPCA模型对影响混输管道内壁均匀腐蚀的影响因素进行分析发现,介质中的H_(2)S含量、CO_(2)含量、温度、流速、pH值对于均匀腐蚀速率的影响相对较大,使用本次研究所提出模型进行均匀腐蚀速率预测的最小误差为1.24%,均方根误差为0.9339%,希尔不等系数为0.5273%。与其它模型相比,本次研究所提模型的精度相对较高,证明使用KPCA-ICS-ELM模型对油气水混输管道进行均匀腐蚀速率预测的可行性及先进性相对较强。Due to the high content of corrosive substances in the oil-gas-water mixed pipeline, the corrosion rate is relatively fast. This research proposes a corrosion rate prediction model based on KPCA-ICS-ELM model. First, the KPCA(nuclear principal component analysis) algorithm, ICS(improved cuckoo search) algorithm, and ELM(extreme learning machine) algorithm are introduced respectively. The KPCA-ICS-ELM model combination method and model evaluation method was proposed. On this basis, the corrosion rate data of the oil-gas-water mixed transportation pipeline was obtained through experiments, and then the index system was constructed. The feasibility of the proposed model is compared with other models to prove the advanced nature of this model wasproposed. The research shows that by using the KPCA model to analyze the influencing factors that affect the uniform corrosion of the inner wall of the mixed transportation pipeline, it is found that the H_(2)S content, CO_(2)content, temperature, flow rate, and pH value in the medium have a relatively large impact on the uniform corrosion rate. The minimum error of the model proposed by the research institute for uniform corrosion rate prediction is 1.24%, the root mean square error is 0.9339%, and the Hill variability coefficient is 0.5273%. Compared with other models, the accuracy of the model proposed in this research is relatively high, which proves that the use of the KPCA-ICS-ELM model to predict the uniform corrosion rate of oil-gas-water mixed pipelines is relatively feasible and advanced.

关 键 词：KPCA-ICS-ELM模型 油气水混输管道 内壁腐蚀 均匀腐蚀 腐蚀速率预测 

分 类 号：TG172.63[金属学及工艺—金属表面处理]