基于差分量子粒子群算法的锅炉NO_x排放模型优化  

Model Improvement for Boiler NO_x Emission Based on DEQPSO Algorithm

作　　者：董泽 马宁[1,2] 孟磊 DONG Ze;MA Ning;MENG Lei(Hebei Engineering Research Center of Simulation & Optimized Control for Power Generation,North China Electric Power University,Baoding 071003,Hebei Province,China;School of Control and Computer Engineering,North China Electric Power University,Beijing102206,China;Datang Environment Industry Group Co.,Ltd,Beijing 100192,China)

机构地区：[1]华北电力大学河北省发电过程仿真与优化控制工程技术研究中心,河北保定071003 [2]华北电力大学控制与计算机工程学院,北京102206 [3]大唐环境产业集团股份有限公司,北京100192

出　　处：《动力工程学报》2019年第3期191-197,共7页Journal of Chinese Society of Power Engineering

基　　金：国家自然科学基金资助项目(2015BJ0030);山西省煤基重点科技攻关资助项目(MD2014-06-06-02)

摘　　要：提出一种基于改进的差分量子粒子群(DEQPSO)算法,将其与超限学习机(ELM)相结合,以某1 000 MW超超临界机组锅炉燃烧系统为研究对象,建立了NO_x排放模型,采用现场样本数据测试所建模型的预测能力,并将该模型的预测结果与基本超限学习机以及引力搜索算法(GSA)、粒子群算法(PSO)和量子粒子群算法(QPSO)优化的超限学习机模型的预测结果进行了对比。结果表明:DEQPSO算法具有更好的参数优化性能,DEQPSO-ELM模型具有较强的泛化能力和良好的预测精度,为电站锅炉NO_x排放质量浓度预测提供了一种有效方法。Taking the boiler combustion system of a 1 000 MW ultra supercritical unit as an object of study,a NOx emission prediction model was established based on the combination of differential evolution quantum particle swarm optimization(DEQPSO)and extreme learning machine(ELM),of which the prediction ability was tested with field sample data,while the prediction results were successively compared with those of basic ELM models,and the ELM models optimized by gravitation search algorithm(GSA),particle swarm optimization(PSO)and quantum particle swarm optimization(QPSO),respectively.Results show that the DEQPSO algorithm has a strong capability in parameter optimization,and the DEQPSO-ELM model has strong generalization ability and high prediction accuracy,which may serve as a reference for NOx emission prediction of power plant boilers.

关 键 词：电站锅炉 NOx 差分量子粒子群算法 超限学习机 

分 类 号：TK22[动力工程及工程热物理—动力机械及工程]