摘要
基于人工神经网络,结合某300MW机组锅炉,通过引入清洁系数F表征锅炉受热面的积灰结渣状况,并针对不同吸热方式的受热面,如高、低温对流受热面,水冷壁等分别讨论了清洁系数的计算方法,建立了受热面污染在线监测模型以及在线数据预处理模型。对低温过热器和RN01吹灰器附近水冷壁区域的灰污曲线实际分析表明,吹灰后受热面清洁因子有较大增加,但吹灰停止后随着运行时间的延长,受热面污染的增加,清洁因子又逐渐恢复至正常值。受热面清洁系数的变化与锅炉实际吹灰情况相吻合,表明该污染监测模型在一定程度上能够正确反映受热面的实际污染状况。
Aiming at optimizing the soot blowing for boiler,pollution status on-line monitoring model for heating surface was established by taking a 300 MW unit boiler as the object.Based on artificial neutral network(ANN),cleaness coefficient F was introduced to represent the fouling condition of heating surface,and the calculation methods of F was discussed respectively,according to different surfaces such as high/low temperature convection heating surface and water wall.A parameter diagnosis model was established for sample data preprocessing to remove 'bad values' and ensure the reality and continuity of thermal parameters.Analysis on ash accumulation curves of water wall area around RN01 soot blower and the low temperature superheater indicates that,the variation of cleaness coefficient is in accordance with the actual soot blowing situation.After soot blowing,the F increases dramatically.But as the operation time extending without soot blowing,it drops back to the normal value.This online model can reflect the actual pollution status of heating surface,to a certain degree.
出处
《热力发电》
CAS
北大核心
2012年第9期99-102,79,共5页
Thermal Power Generation
基金
中央高校基本科研业务费专项资金资助(09ZG02)
关键词
300
MW机组
锅炉
对流受热面
水冷壁
结渣
吹灰
清洁因子
监测模型
boiler
artificial neural network
convection heating surface
water wall
fouling and slagging
on-line monitoring
