摘要
腐蚀是造成油气管道失效的主要原因,准确预测管道的腐蚀缺陷是防止管道失效事故的重要手段。基于斯皮尔曼(Spearman)、布谷鸟算法(CS)和极限学习机(ELM)组合模型,采用Spearman相关系数判别腐蚀因素的相关性,利用因子分析进行降维处理,引入极限学习机对腐蚀速率进行回归,采用CS算法对ELM模型的输入权值和隐含层阈值进行迭代寻优,并比较不同的ELM激活函数,建立了一套埋地管道腐蚀速率预测方法。通过对某埋地管道进行预测值与实际检测值的比对,腐蚀速率的平均相对误差为2.32%。
Corrosion is the main cause of oil and gas pipeline failure.Accurately predicting corrosion defects is an important means to prevent pipeline failure accidents.This work is based on the combination model of Spearman,Cuckoo algorithm(CS)and Extreme Learning Machine(ELM).The Spearman correlation coefficient is used to distinguish the correlation of corrosion factors,factor analysis is used to reduce dimension,and Extreme Learning Machineis introduced to regression the corrosion rate.The CS algorithm is used to iteratively optimize the input weights and hidden layer thresholds of the ELM model,and different ELM activation functions are compared to establish a set of prediction methods for the corrosion rate of buried pipelines.By comparing the predicted value and the actual measured value of a buried pipeline,the average relative error of the corrosion rate is 2.32%.
作者
李世强
杨国栋
金龙
马宁
郄晓敏
王春洁
LI Shiqiang;YANG Guodong;JIN Long;MA Ning;QIE Xiaomin;WANG Chunjie(No.4 Oil Production Plant of Huabei Oilfield Company,CNPC;Hebei Huabei Petroleum Engineering Construction Co.,Ltd.;Gas Store Management Agency of Huabei Oilfield Company;Erlian Filiale of HuaBei Oilfield Company,CNPC)
出处
《油气田地面工程》
2022年第12期17-22,共6页
Oil-Gas Field Surface Engineering
