摘要
在延迟焦化装置中,碳钢加热炉存在燃烧不稳定性,易造成加热炉管内局部超温而损坏的问题,因此必须实时检测加热炉内各个部位的温度。针对这一问题,拟开展基于卷积长短时记忆神经网络的炉温在线预报方法研究,通过对高温炉温变化规律的分析,实现对高温炉温变化的预测,并将平均误差控制在31.5 Kalvin以下。
In delayed coking equipment,carbon steel heating furnaces have combustion instability,which can easily cause local overheating and damage in the heating furnace tubes.Therefore,it is necessary to monitor the temperature of various parts of the heating furnace in real-time.In response to this issue,it is planned to conduct research on the online prediction method of furnace temperature based on convolutional long and short time memory neural networks.By analyzing the changes in high-temperature furnace temperature,the prediction of high-temperature furnace temperature changes is achieved,and the average error is controlled below 31.5 Kalvin.
作者
李朝阳
Li Chaoyang(Shanxin Software Rizhao Branch,Rizhao Shandong 276800,China)
出处
《山西冶金》
CAS
2023年第8期125-126,132,共3页
Shanxi Metallurgy
关键词
加热炉管道
神经网络
实时温度
损坏检测
heating furnace pipeline
neural network
teal time temperature
damage detection
