基于云模型和DSmT的风电机组状态评估方法  

Wind Turbine Condition Assessment Method Based on Cloud Model and Dezert-Smarandache Theory

作　　者：刘林 刘沙 刘亚 陈俊生 LIU Lin;LIU Sha;LIU Ya;CHEN Junsheng(Southern Offshore Wind Power Joint Development Co.,Ltd.,Zhuhai,Guangdong 519080,China;School of Electrical Engineering,Chongqing University,Chongqing 400065,China;College of Automation,Chongqing University of Posts and Telecommunications,Chongqing 400000,China)

机构地区：[1]南方海上风电联合开发有限公司,广东珠海519080 [2]重庆大学电气工程学院,重庆400065 [3]重庆邮电大学自动化学院,重庆400000

出　　处：《广东电力》2021年第2期45-53,共9页Guangdong Electric Power

基　　金：重庆市自然科学基金面上项目(cstc2020jcyj-msxmX0687)。

摘　　要：为准确掌握风电机组的实际运行状态,需要利用已有的运行数据对风电机组运行状态进行研究,从而为其他风电机组的安全经济运行提供依据。根据风电机组监控与数据采集(supervisory control and data acquisition,SCADA)系统和风电机组状态监测系统(condition monitoring system,CMS)获得的监测数据,提出基于云模型和Dezert-Smarandache理论(DSmT)的风电机组状态评估方法。首先确定影响风电机组项目层的评价指标,建立多源参数融合的两级状态评价指标体系;其次确定各评估指标的动态劣化度,建立云模型,求出隶属度;再根据隶属度求得概率质量(mass)函数,采用DsmT对mass函数进行融合,在融合过程中引入mass函数的快速收敛算法以减少计算量,按照最大信任规则,确定评估状态;最后以我国华东地区某风场的风电机组健康状态评估为例对所提方法的有效性进行验证。与当前的风电机组评估方法对比,所提风电机组评估方法更加准确。In order to accurately grasp the actual operating state of the wind turbine generator,it is necessary to use the existing operating data to study the operating state of the wind turbine generator,so as to provide a basis for the safe and economic operation of other wind turbines.According to monitoring data of the supervisory control and data acquisition(SCADA)system and the condition monitoring system(CMS),this paper proposes an assessment method based on the cloud model and DSmT for the wind turbines.It firstlt determines the assessment indicators affecting the project level of the wind turbine generator,and establishes a two-level state assessment index system for multi-source parameter fusion.Secondly,it confirms the dynamic degradation degree of the assessment index,builds the cloud model and calculates the membership degree.Afterwards,it calculates the mass function according to the membership degree,and uses Dezert-Smarandache theory(DsmT)for fusion of the mass function.By introducing a rapid convergence algorithm in the fusion process,it is able to reduce calculation amount and according to the maximum trust rule,it is able to determine the assessment state.Finally,the effectiveness of the proposed method is verified by taking the wind turbine health status assessment of a wind farm in East China as an example.The paper proves compared with current assessment methods for the wind turbines,the proposed method is more accurate.

关 键 词：风电机组 状态评估 动态劣化度 云模型 DSMT 

分 类 号：TM614[电气工程—电力系统及自动化]