王伟,常浩,王宝玉

• 1:华电电力科学研究院

摘要(Abstract)：

针对锅炉飞灰含碳量难以长期准确预测的问题,从提高模型预测精度和自适应能力的角度出发,提出一种基于模型预测性能评价的自适应校正加权最小二乘支持向量机(WLSSVM)软测量模型。构造了基于最大线性无关组的软测量模型训练样本集,使WLSSVM模型具有较好的稀疏性,并减少了训练过程的计算量;建立基于数据相似度加权因子的WLSSVM软测量模型,利用双种群差分进化算法进行模型参数的优化选取;通过模型预测性能在线评估和递推校正实现了模型在线自适应校正。在某台300MW机组锅炉上进行的仿真试验结果表明,该算法模型具有良好的预测精度和自适应能力,能够有效预测锅炉飞灰含碳量。

关键词(KeyWords)： 锅炉;飞灰;含碳量;最大线性无关组;双种群差分进化算法;递推校正

Abstract:

Keywords:

基金项目(Foundation):

作者(Author): 王伟,常浩,王宝玉

参考文献(References)：

• [1]潘理黎,王佳莹,杨玉峰,等.火电厂飞灰含碳量在线监测设备现状[J].热力发电,2008(11):10-14.PAN Lili,WANG Jiaying,YANG Yufeng,et al.Statusquo of the on-line monitors for carbon content in flyash at coal-fired power plants[J].Thermal Power Gen-eration,2008(11):10-14.
• [2]陈敏生,刘定平.基于核主元分分析和支持向量机的电站锅炉飞灰含碳量软测量建模[J].华北电力大学学报,2006,33(1):72-75.CHEN Minsheng,LIU Dingping.Soft-sensing model-ing of the unburned carbon in fly ash based on KPCA-SVM for power station boilers[J].Journal of NorthChina Electric Power University,2006,33(1):72-75.
• [3]韩璞,乔弘,翟永杰,等.基于支持向量机融合的多模型动态飞灰含碳量软测量[C]∥第20届中国控制与决策会议论文集.烟台:[出版者不详],2008:4595-4599.HAN Pu,QIAO Hong,ZHAI Yongjie,et al.Soft-sens-ing for multiple models of carbon content of fly ashbased on SVM fusion theory[C]∥Proceedings of the20th Chinese Control and Decision Conference.Yantai,2008:4595-4599.
• [4]李钧,阎维平,李春,等.基于预数值计算的锅炉飞灰可燃物含量建模[J].中国电机工程学报,2009,29(17):32-37.LI Jun,YAN Weiping,LI Chun,et al.Modeling of theunburned carbon in fly ash based on numerical simula-tion in the utility boiler[J].Proceedings of the CSEE,2009,29(17):32-37.
• [5]张贵炜,鲍琳,李奇伟.基于信息融合的火电厂飞灰含碳量的软测量建模[J].信息与控制,2009,38(6):646-652.ZHANG Guiwei,BAO Lin,LI Qiwei.Soft-sensingmodeling of the carbon content in fly ash based on in-formation fusion for power plant[J].Information andControl,2009,38(6):646-652.
• [6]LIU C,GAO Y,HOU X,et al.Optimized control onmodel of unburned carbon content in fly ash of stationboilers[C]∥Proceedings of Sixth International Confer-ence on Natural Computation.Baoding,2010:903-908.
• [7]温文杰,马晓茜,刘翱.锅炉混煤掺烧的飞灰含碳量预测与运行优化[J].热力发电,2010(3):30-35.WEN Wenjie,MA Xiaoqian,LIU Ao.Prediction of un-burned carbon content in fly ash and operation optimi-zation for mixedly burning blended coal in boilers[J].Thermal Power Generation,2010(3):30-35.
• [8]顾燕萍,赵文杰,吴占松.最小二乘支持向量机的算法研究[J].清华大学学报(自然科学版),2010,50(7):1063-1066.GU Yanping,ZHAO Wenjie,WU Zhansong.Leastsquares support vector machine algorithm[J].Journalof Tsinghua University(Science and Technology),2010,50(7):1063-1066.
• [9]Suykens J A K,Brabeanter De J,Lukas L,et al.Weigh-ted least squares support vector machines:robustnessand sparse approximation[J].Neurocomputing,2002,48(1-4):85-105.
• [10]司刚全,曹晖,张彦斌,等.一种基于密度加权的最小二乘支持向量机稀疏化算法[J].西安交通大学学报,2009,43(10):11-15.SI Gangquan,CAO Hui,ZHANG Yanbin,et al.Densi-ty weighted pruning method for sparse least squaressupport vector machines[J].Journal of Xi'an JiaotongUniversity,2009,43(10):11-15.