基于数据挖掘技术的锅炉岛运行能效诊断

Energy Efficiency Diagnosis of Boiler Island Operation Based on Data Mining Technology

为应对火电机组深度调峰背景下锅炉岛运行条件多变时能效诊断问题,提出基于数据挖掘技术构建能效基准库,结合耗差因子理论建立锅炉岛的运行能效诊断模型,实现对不同运行条件下锅炉岛的能效诊断。将负荷、煤质及环境温度作为锅炉岛运行外部约束条件,通过改进的K-means算法对历史运行数据进行工况划分;结合灰色关联分析与K近邻算法构造了工况相似度计算模型,以相似工况中能效状态最优的工况构建了能效基准工况库;通过分析运行工况与基准工况的偏离程度实现了对锅炉岛运行能效的评价,并对基准工况库进行更新,基于能效特征参数的煤耗偏差及耗差因子实现对锅炉岛运行能效逐层分析,从而得到造成能效偏离的具体原因,便于运行人员及时做出运行调整。

In order to solve the problem of energy efficiency diagnosis of boiler island under the background of thermal power unit peak load adjustment,this paper constructed the energy efficiency reference base based on data mining technology,and established the energy efficiency diagnosis model of boiler island under different operating conditions combined with the consumption difference factor theory.Taking the load,coal quality and ambient temperature as the external constraints of boiler island operation,the historical operation data is divided into working conditions by the improved K-means algorithm.The operating conditions with the best energy efficiency state among similar operating conditions construct the energy efficiency benchmark condition library;By analyzing the deviation degree between the operating conditions and the benchmark condition,the evaluation of the operating energy efficiency of the boiler island is realized,and the benchmark condition library is updated.Based on the coal consumption deviation and consumption difference factor of energy efficiency characteristic parameters,the energy efficiency of boiler island operation can be analyzed layer by layer,so as to obtain the specific reasons for the deviation of energy efficiency,which is convenient for operators to make timely operation adjustments.