基于工况划分的电厂燃煤锅炉多目标模糊优化方法  

Multi-objective fuzzy optimization method for coal-fired boilers in power plant based on working conditions

作　　者：曹歌瀚 黄亚继[1] 陈波 张荣初 刘宇青 邹怡然 CAO Gehan;HUANG Yaji;CHEN Bo;ZHANG Rongchu;LIU Yuqing;ZOU Yiran(Key Laboratory of Energy Thermal Conversion and Control of Ministry of Education,Southeast University,Nanjing 210096,China;Jiangsu Frontier Electric Power Technology Co.,Ltd.,Nanjing 211102,China;Nanjing Changrong Acoustic Inc.,Nanjing 210008,China)

机构地区：[1]东南大学能源热转换及其过程测控教育部重点实验室,江苏南京210096 [2]江苏方天电力技术有限公司,江苏南京211102 [3]南京常荣声学股份有限公司,江苏南京210008

出　　处：《热科学与技术》2025年第1期65-71,共7页Journal of Thermal Science and Technology

基　　金：江苏省科技成果转化专项资金资助项目(BA2020001)。

摘　　要：为了综合优化电厂燃煤锅炉经济性与环保性,利用锅炉原有的历史数据指导锅炉运行十分重要。借助于电厂的历史数据库,结合数据挖掘算法对某电厂燃煤锅炉进行多目标优化。首先,利用K均值聚类算法对锅炉工况进行划分,通过误差平方和曲线的曲率变化确定工况数。其次,利用多目标模糊优化方法综合考虑锅炉效率与炉膛NOx生成量的优化,构建模糊隶属函数并求解得出当前工况的最佳锅炉运行参数。计算结果显示:K均值聚类方法将锅炉运行工况按负荷分为4类,按煤质分为3类。多目标模糊优化结果显示:在60.00%负荷、煤质中等的工况下,相较于只考虑锅炉效率为最大时的单目标优化结果,综合考虑锅炉效率和炉膛NO_(x)生成量的结果使得锅炉效率仅仅下降1.30%的同时炉膛NOx生成量下降了31.18%,工况满意度达到了0.9212,给出了该工况下锅炉效率与炉膛NOx生成量取得相对平衡的解决方案。In order to comprehensively optimize the economy and environmental performance of coal-fired boilers in power plants,it is very important to use the original historical data of the boilers to guide their operation.With the help of the historical database of power plant,combined with the data mining algorithm,the multi-objective optimization for the coal-fired boiler in a power plant is carried out.Firstly,the K-means clustering algorithm is used to divide the boiler working conditions,and the number of working conditions is determined by the curvature change in the error sum of squares curve.Secondly,a multi-objective fuzzy optimization method is used to simultaneously optimize boiler efficiency and NO,emission.A fuzzy membership function is constructed and the optimal boiler operating parameters are determined for the current operating conditions.The results show that the K-means clustering method divides the boiler operating conditions into 4 categories according to load and 3 categories according to the coal quality.The multi-objective fuzzy optimization results show that at a 60.00% load and medium coal quality,compared to the single-objective optimization focusing solely on maximizing boiler efficiency,the multi-objective optimization approach results in only a 1.30% reduction in boiler efficiency,while the NO_(x) emissions decrease by 31.18%.And the working condition satisfaction reaches 0.9212.The solution offers a balanced approach to optimizing both boiler efficiency and NO,emission under these working conditionsen.

关 键 词：燃煤锅炉 数据挖掘 工况划分 K均值聚类 多目标模糊优化 

分 类 号：TK229.6[动力工程及工程热物理—动力机械及工程]