摘要
气化炉内是高温(一般均超过1 050℃)、高压(约6.0 MPa)、强腐蚀环境,并伴随着高强度的气流冲刷,使得测温元件高温热电偶的工作寿命很短,无法对气化炉炉膛温度进行实时监测,导致操作的滞后对气化炉以及生产系统会产生巨大的影响,甚至造成气化炉渣堵系统停车。本文通过变量选择、数据采集与处理并采用BP神经网络法建立气化炉炉膛温度软测量模型,并进行效果验证,通过误差分析,表明基于BP神经网络法建立的炉膛温度软测量模型,能够有效的指导气化操作和化工生产,为今后气化炉炉膛温度软测量的发展做出一定的指导和建议。
The gasifier is high temperature(generally more than 1050T),high pressure(about 6.0MPa),strong corrosive environment,and accompanied hy high-strength air flow erosion,which makes working life of the high-temperature thermocouple of temperature measuring element very short and the furnace temperature of gasifier cannot be monitored in real time,resulting in the lagging of operation that will have a huge impact on the temperature of gasifier and production system,and even cause the gasification slag plugging system to shut down.In this paper,through variable selection,data acquisition and processing,and using BP neural network method to establish a gasifier furnace temperature soft sensor model,and to verify the effect,through error analysis,it is showed that the furnace temperature soft sensor model based on BP neural network method can effectively guide the gasification operation and chemical production,and make certain guidance and suggestions for the future development of gasifier furnace temperature soft sensor.
作者
李天伦
LI Le-lun(Yankuang Xinjiang Coal Chemical Co.,Ltd.,Urumqi 830000,Xinjiang Province)
出处
《氮肥技术》
2020年第3期16-21,共6页
Nitrogenous Fertilizer Technology
关键词
神经网络
软测量
问题
建模
指标
效果
neural network
soft sensor
problem
modeling
index
effect
