基于DPC-MND和多元状态估计的磨煤机故障预警研究  

Research on Fault Warning of Coal Mills Based on DPC-MND and Multivariate State Estimation Technique

作　　者：姚天杨 茅大钧 韩万里 YAO Tianyang;MAO Dajun;HAN WanLi(School of Automation Engineering,Shanghai University of Electric Power,Shanghai 200090,China;State Grid Shanghai Electric Power Company Qingpu Power Supply Company,Shanghai 201700,China)

机构地区：[1]上海电力大学自动化工程学院,上海200090 [2]国网上海市电力公司青浦供电公司,上海201700

出　　处：《控制工程》2024年第5期928-937,共10页Control Engineering of China

基　　金：上海市“科技创新行动计划”地方院校建设专项资助项目(19020500700);中国华能集团有限公司2020年度科技项目(HNKJ-F2002);中国华电集团有限公司2017年度科技攻关项目(CHDJSKJ17-01-07)。

摘　　要：火电机组磨煤机存在运行条件恶劣、故障频发等问题,对磨煤机进行故障预警,可以有效防止一些常见故障的发生,从而保证火电机组的安全运行。为此,提出一种基于相互邻近度的密度峰值聚类和多元状态估计的磨煤机故障预警方法。首先,采用核主元分析选取磨煤机的主要状态参数,同时采用集合经验模态分解对历史运行数据进行去噪,进一步优化数据质量;然后,采用基于相互邻近度的密度峰值聚类(density peaks clustering based on mutual neighborhood degrees,DPC-MND)方法构建动态记忆矩阵,利用多元状态估计技术(multivariate state estimation techniques,MSET)对磨煤机正常运行工况下的历史数据进行建模,并确定磨煤机的运行状态。最后,以安徽某电厂ZGM113G型中速磨煤机为例进行验证,结果表明该方法可以实现对磨煤机故障的有效预警。Coal mill of thermal power units have problems such as harsh operating conditions and frequent failures.Early warning of coal mill can effectively prevent the occurrence of some common faults,thereby ensuring the safe operation of thermal power units.Therefore,the density peaks clustering based on mutual neighborhood degrees and multivariate state estimation techniques are proposed for early warning of coal mill faults.Firstly,kernel principal component analysis is used to select the main state parameters of the coal mill.At the same time,ensemble empirical mode decomposition is adopted to denoise the data and further optimize data quality.Then,the density peaks clustering based on mutual neighborhood degrees(DPC-MND)method is applied to construct a dynamic memory matrix and a prediction model is established by historical data under normal operating conditions of the coal mill through multivariate state estimation techniques(MSET),and determines operating status of the coal mill.Finally,a ZGM113G medium-speed coal mill in a power plant in Anhui is used as an example to verify.The results indicates that this method can achieve effective early warning of coal mill failures.

关 键 词：中速磨煤机 核主元分析 DPC-MND 多元状态估计技术 故障预警 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置]