一种改进型LSSVM模型在电站锅炉燃烧与优化中的应用  

Application of an Improved LSSVM in Combustion Modeling and Optimization of Utility Boilers

作　　者：李应保 王东风 LI Yingbao;WANG Dongfeng(Department of Automation, North China Electric Power University, Baoding 071003, Hebei Province, China;Hebei Engineering Research Center of Simulation ＆ Optimized Control for Power Generation, North China Electric Power University, Baoding 071003, Hebei Province, China)

机构地区：[1]华北电力大学自动化系,河北保定071003 [2]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

出　　处：《动力工程学报》2018年第4期258-264,共7页Journal of Chinese Society of Power Engineering

基　　金：国家自然科学基金资助项目(61203041);中央高校基本科研业务费专项资金资助项目(20140139);河北省自然科学基金资助项目(F2014502059)

摘　　要：采用鲁棒交叉验证算法对传统最小二乘支持向量机模型进行改进,并对某350 MW机组的锅炉燃烧系统进行建模,用以预测锅炉出口烟气参数的变化。结果表明:鲁棒交叉验证最小二乘支持向量机建模方法提高了模型的预测精度,增强了模型的鲁棒性,有效避免了单一元素误差对预测结果的影响;通过优化锅炉燃烧系统输入参数的组合,可以有效提高锅炉的整体经济效益并降低NOx的生成量。The traditional least squares support vector machine（LSSVM）model was improved by robust cross-validation（RCV）method,while the combustion system of a 350 MW domestic boiler was modeled to predict the change of outlet flue gas parameters of the boiler.Results show that the robust cross-validation least squares support vector machine（RCV-LSSVM）modeling method improves the predicting precision and enhances the robustness of the model,which can effectively avoid the influence of single element error upon the model.By optimizing the combination of input parameters in the boiler combustion system,the overall economic efficiency of boiler can be improved,while the generation of nitrogen oxides（NOx）can be reduced.

关 键 词：燃烧优化 预测模型 鲁棒交叉验证 最小二乘支持向量机 经济效益 NOx 

分 类 号：TK227.1[动力工程及工程热物理—动力机械及工程]