基于Attention-CS-LSTM乙烯裂解炉管温度预测  

Prediction of Temperature Field in an Improved CS-LSTM Ethylene Cracking Furnace

作　　者：张子默 崔得龙 ZHANG Zimo;CUI Delong(jilin Institute of Chemical Technology School of Infomation and Control Engineering 132022,China;school of Electronic Infomation Engineering,Guangdong University of Petrochemical Technology 525000,China)

机构地区：[1]吉林化工学院信息与控制工程学院,吉林吉林132022 [2]广东石油化工学院计算机与电子信息学院,广东茂名525000

出　　处：《长江信息通信》2024年第4期43-46,共4页Changjiang Information & Communications

摘　　要：在乙烯生产过程中,针对乙烯裂解炉管温度难监测的情况,需要对传统的温度测量方法进行改进,通过数据模型下的优化操作可以有效预测乙烯裂解炉出口温度,当出现温度波动时进行干预,提高产品效率和生产安全。文章将改进的布谷鸟算法优化LSTM(CS-LSTM)应用于真实工业数据,并与四种模型进行比较。仿真结果表明,采用Attention-CS-LSTM预测准确率明显提高,且具有良好的稳态准确度,该方法的温度预测准确率为95%。In the process of ethylene production,it is necessary to improve the traditional temperature measurement method in response to the difficulty of monitoring the temperature of the cthylene cracking furnace tube.By optimizing the operation under the data model,the outlet temperature of the ethylene cracking furnace can be effectively predicted.When temperature fluctuations occur,intervention can be taken to improve product efficiency and production safety.This article applies the improved cuckoo bird algorithm optimized LSTM(CS-LSTM)to real industrial data and compares it with four models.The simulation results show that using Attention CS LSTM significantly improves the prediction accuracy and has good steady-state accuracy,with a temperature prediction accuracy of 95%.

关 键 词：布谷鸟算法 LSTM 注意力机制 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]