摘要
针对液化天然气(LNG)储罐的腐蚀问题,提出基于灰狼优化最小二乘支持向量机(GWOLSSVM)的LNG储罐腐蚀速率预测组合算法,以我国某LNG储罐储罐为例,使用该种组合算法对罐底、罐壁以及罐顶等位置的腐蚀速率进行预测。研究结果表明:使用经过交叉验证以后的GWO-LSSVM组合模型对LNG储罐罐底腐蚀速率进行预测的平均相对误差为2.07%,对罐壁进行腐蚀速率预测的平均相对误差为4.30%,对罐顶进行腐蚀速率预测的平均相对误差为3.10%,3个位置的预测结果均好于其他预测模型,表明使用该种类型的算法可以对LNG储罐不同位置的腐蚀速率进行有效的预测。
Aiming at the corrosion problem of liquefied natural gas(LNG)storage tanks,a combined algorithm for predicting the corrosion rate of LNG storage tank based on Grey Wolf Optimized Least Squares Support Vector Machine(GWO-LSSVM)is proposed.Taking a certain LNG storage tank as an example,the combined algorithm is used to predict the corrosion rate of the bottom,wall and top of the tank.The results show that the average relative error of corrosion rate prediction of LNG storage tank bottom using the cross-validated GWO-LSSVM model is 2.07%,the average relative error of corrosion rate prediction for tank wall is 4.30%,and the average relative error of corrosion rate prediction for tank top is 3.10%,the prediction results of the three locations are better than other prediction models,which prove that the use of this type of algorithm can effectively predict the corrosion rate at different locations of LNG storage tanks.
作者
郭海新
GUO Haixin(Kunlun Energy Hubei Huanggang Liquefied Natural Gas Co.,Ltd.,Huanggang 438000,China)
出处
《能源化工》
CAS
2021年第4期61-67,共7页
Energy Chemical Industry
