测控技术2025，Vol.44Issue(2)：53-59,67,8.DOI:10.19708/j.ckjs.2025.02.304
用于天然气管网腐蚀速率预测的WT-LSTM网络
WT-LSTM Network for Corrosion Rate Prediction in Natural Gas Pipeline Networks
刘晓莉 1何健聪2
作者信息
- 1. 天然气净化总厂信息科技部,重庆 400000
- 2. 湖南省鹰眼在线电子科技有限公司,湖南长沙 430100
- 折叠
摘要
Abstract
During the natural gas purification process,strong corrosive gases lead to varying degrees of attrition of the reaction vessel.Although traditional corrosion probes can provide corrosion warning,they have a certain lag.To solve this problem,a corrosion data trend prediction method based on wavelet transform-long short term memory(WT-LSTM)network is proposed.The original data are decomposed using db10 wavelet,and after ob-taining each detail component,the LSTM network is used to predict each component and reconstruct the signal.The experimental results show that the root mean square error(RMSE)of the WT-LSTM model is 0.002 562,which is lower than the RMSE value of 0.003 178 for the LSTM-only model,indicating that the WT-LSTM net-work is more accurate in trend prediction.The prediction scheme based on WT-LSTM network can effectively track the changes of corrosion data,especially the effect is significant when the data change abruptly,so as to enhance the online monitoring capability of the corrosion probe,realize the prediction and warning of corrosion,and ensure the stable operation of the natural gas purification process.关键词
天然气净化/腐蚀探针/WT-LSTM网络/db10小波/预警/预测Key words
natural gas purification/corrosion probes/WT-LSTM network/db10 wavelet/warning/prediction分类
石油、天然气工程引用本文复制引用
刘晓莉,何健聪..用于天然气管网腐蚀速率预测的WT-LSTM网络[J].测控技术,2025,44(2):53-59,67,8.
