基于神经网络的GE水煤浆气化炉炉温预测研究  

GE Coal Water Slurry Gasifier Temperature Prediction Based on Neural Network

作　　者：邓丽君 Deng Lijun(Sinopec Ningbo Engineering Co.Ltd.,Ningbo,315103,China)

机构地区：[1]中石化宁波工程有限公司,浙江宁波315103

出　　处：《石油化工自动化》2024年第6期24-28,共5页Automation in Petro-chemical Industry

摘　　要：GE水煤浆气化炉炉温的控制难度较高,若调节不及时易导致气化炉超温,对气化炉炉体、耐火砖损耗、合成气品质及后续生产流程造成诸多危害;气化炉炉膛热电偶会频繁损坏或漂移,将导致操作工无法正确判断正常炉温。针对GE水煤浆气化炉温度建立一种基于BP神经网络模型的预测方法,根据模型现有决定因素预测气化炉炉温,及时调整气化炉温度尽量避免或降低炉温超温带来的损失,同时减少频繁更换炉膛热电偶引起的经济损失。The control of the furnace temperature of GE coal water slurry gasifier is relatively difficulty.If not adjusted in time,overtemperature of the gasifier is easy,which can cause many hazards to the gasifier body,refractory brick loss,synthesis gas quality,and subsequent production processes.The thermocouple in the gasifier furnace will be frequently damaged or drift,which will cause the incorrect judgement of normal furnace temperature to the operator.Therefore,a prediction method of GE coal-water slurry gasifier temperature based on BP neural network model is established,neural network the temperature of gasifier can be predicteel and adjusted in time according to the existing determinants of the model,the loss caused by overtemperature of gasifier can be avoided or reduced as far as possible,the economic loss caused by frequent replacement of thermocouples of gasifier can be reduced.

关 键 词：气化炉炉温 BP神经网络 预测拟合 影响权重 MATLAB软件 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]