摘要
当前,我国电力行业仍以煤炭为主要能源消耗品,加之电站锅炉常用燃煤中的硫含量与灰分较高,易造成受热面的积灰和结渣,而炉膛严重结渣将限制锅炉出力,威胁机组运行的经济性与安全性,因此,开发一种全面、综合的结渣预测模型对锅炉炉膛结渣程度有效监测非常关键。将模糊数学理论与BP神经网络相结合,构建适用于燃煤电站锅炉炉膛结渣特性判定的模糊神经网络。在选择输入评判指标时,从煤灰本身考虑其结渣特性,同时将无因次炉膛最高温度纳入模型,考虑锅炉运行情况,使判别依据更加全面,共选取了分辨率较高且最具代表性的6个因素作为本模型的输入判别指标。采用4种不同类型的隶属函数,将判别指标模糊化后,作为模糊神经网络的模型输入,并与不进行模糊化处理的神经网络对比,根据统计学原理,选用出现概率最大的结果作为最终评判指标,增加预测结果的精确度。针对华能秦岭电厂660 MW超临界锅炉BMCR负荷运行时炉膛结渣情况,采用构造好的炉膛结渣模糊神经网络模型进行预测。结果表明,该机组在燃用华亭煤时,炉膛结渣判别指标软化温度、硅比、硅铝比、碱酸比、综合指标、炉膛无因次最高温度分别为1220℃、58.71、1.63、0.48、2.55、0.982,为重度结渣。在适当掺烧黄陵1号煤时,各项指标则分别为1255℃、71.02、2.04、0.31、2.15、0.958,为中等结渣,因此可采用掺烧优质煤来改善炉膛结渣状况。该模型预测结果准确,为综合评价锅炉炉膛的结渣特性提供了新方法。
At present,coal is still the main energy consumption in the power industry.In addition,the sulfur content and ash content in the commonly used coal for power station boilers is high,which is easy to cause ash and slagging on the heated surface.The serious slag⁃ging in the furnace will limit the output of the boiler and threaten the economy and safety of the unit operation,therefore,the development of a comprehensive and comprehensive slagging prediction model will be the focus of further research,which is very important to effectively monitor the degree of slagging in boiler furnace and its development trend.Combining fuzzy mathematics theory with BP neural network,a fuzzy neural network suitable for judging characteristics of slagging in the furnace of coal-fired power plant was constructed.When selec⁃ting the input evaluation index,not only its slagging characteristics from the coal ash itself were considered,but also the dimensionless fur⁃nace maximum temperature,which reflects the slagging judgment index of unit operation,was incorporated into the model.Taking the oper⁃ating conditions of the boiler into account,the judgment basis is more comprehensive.A total of 6 factors with higher resolution and the most representative were selected as the input discriminant indicators of this model.Four different types of membership functions were used to fuzz the discriminative index as the input of the fuzzy neural network model,and the neural network without fuzzification was used as the comparison.According to the principles of statistics,the result with the highest occurrence probability was selected as the final eval⁃uation index to increase the accuracy of the prediction result.The results show that when the unit burns Huating coal,the furnace slagging discrimination indexes softening temperature,silicon ratio,silicon aluminum ratio,alkali acid ratio,comprehensive index and the dimension⁃less maximum temperature of the furnace are 1220℃,58.71,1.63,0.48,2.55 and 0.982 respectively,which are severe slagging.When Huangling No.1 coal is properly mixed,it is 1255℃,71.02,2.04,0.31,2.15 and 0.958 respectively,which is medium slagging.There⁃fore,proper blending of high-quality coal can be used to improve the slagging condition of the furnace.The prediction result of this model is accurate,which can provides a new way to comprehensively evaluate the slagging characteristics of the boiler furnace.
作者
朱超
郁翔
李峰
周熙宏
毕凌峰
杨冬
ZHU Chao;YU Xiang;LI Feng;ZHOU Xihong;BI Lingfeng;YANG Dong(Electric Power Research Institute of State Grid Shaanxi Electric Power Company,Xi′an 710100,China;State Key Laboatory of Multiphase Flow in Power Engineering,Xi′an Jiaotong University,Xi′an 710049,China)
出处
《洁净煤技术》
CAS
北大核心
2022年第4期175-182,共8页
Clean Coal Technology
基金
国网陕西省电力公司科技资助项目(5226KY19004H)。
关键词
燃煤锅炉
炉膛结渣
模糊理论
神经网络
隶属函数
模式识别
coal-fired boiler
furnace slagging
fuzzy theory
neural network
membership function
pattern recogntion