摘要
目前,燃煤机组广泛采用选择性催化还原(SCR)脱硝技术。对燃煤机组NO_x排放浓度进行准确预测,不但有利于进一步提高SCR控制系统的调节品质,而且可以评估现场所收集到的数据是否真实准确,为环保部门对电厂排放NO_x浓度进行监管执法提供依据。提出一种基于无迹卡尔曼滤波最小二乘支持向量机(UKF-LSSVM)的NO_x排放浓度预测方法。基于现场数据和理论分析,确定了脱硝系统动态模型的输入和输出变量。采用无迹卡尔曼滤波(UKF)不断更新核参数σ和其他模型参数α、b,并采用样本更新策略对支持向量进行更新,提高了模型的自适应能力。将该方法用于某300 MW机组脱硝系统的NO_x排放浓度预测。仿真结果表明,所建模型能够准确预测燃煤机组NO_x排放浓度。与原有基于批量最小二乘支持向量机(LSSVM)建立的稳态模型相比,该方法具有更高的预测精度和自适应能力,为进一步研究脱硝系统动态优化控制奠定了基础。
At present,the selective catalytic reduction(SCR)denitration technology is widely used in coal-fired units.Accurately predicting the NO x emission concentration not only helps to improve the regulation quality of the SCR control system,but also to evaluate the authenticity and accuracy of the data collected from the field to provide evidences for the supervision and enforcement by environmental protection authority for NO x emission concentration of power plants.A prediction method based on least squares support vector machine and unscented Kalman filter(UKF-LSSVM)is proposed to predict the NO x emission concentration of coal-fired units.On the basis of field data and theoretic analysis,input and output variables of dynamics model for denitration system are determined.The kernel parameterσand model parameters(α、b)are continuously updated by using unscented Kalman filter(UKF),and support vectors are updated through sample update strategy,thus the adaptability of the model is improved.This method is applied to the NO x concentration prediction for denitration system in certain300MW unit.The simulation results show that the model can accurately predict the NO x emission concentration of coal-fired units.Compared with the original steady-state model based on batch least squares support vector machine(LSSVM),this method has higher prediction accuracy and self-adaptive ability,which lays a foundation for further research on dynamic optimization control of denitration system.
作者
张友卫
曹硕硕
魏威
李春岩
曾令超
杨晨琛
李益国
ZHANG Youwei;CAO Shuoshuo;WEI Wei;LI Chunyan;ZENG Lingchao;YANG Chenchen;LI Yiguo(Jiangsu Frontier Electric Technology Co. ,Ltd. ,Nanjing 211102,China;School of Energy and Environment,Southeast University,Nanjing 210096,China)
出处
《自动化仪表》
CAS
2018年第12期13-17,共5页
Process Automation Instrumentation
基金
国家自然科学基金资助项目(51476027)
关键词
无迹卡尔曼滤波
最小二乘支持向量机
燃煤机组
脱硝系统
NOx排放浓度
动态建模
自适应
Unscented Kalman filter(UKF)
Least squares support vector machine(LSSVM)
Coal-fired unit
Denitration system
NOx emission concentration
Dynamic modeling
Self- adaptation
作者简介
张友卫(1986-),男,学士,工程师,主要从事火力发电节能减排系统的研究,E-mail:15905166639@163.com;李益国(通信作者),男,博士,教授,主要从事热工自动化的教学和科研工作,E-mail:lyg@seu.edu.cn.
