摘要
为预警风电机组潜在故障、增强机组出力安全性,基于监控和数据采集(SCADA)系统,提出一种异常数据重建的风电机组故障预警方法。首先,充分利用同风场风机SCADA数据,分别重建输入类与输出类目标机组数据,克服了部分数据信息丢失、数据异常问题;其次,使用提取的代表性数据建立故障预警模型,所得预警模型更贴近机组运行动态特性;最后,采用改进的衰退指标预警潜在故障,直观展示机组阶段性衰退程度。案例研究中使用某风电场SCADA故障数据,并使用3种标准确定所提策略参数设定值,结果表明可至少提前3周预警风电机组齿轮箱潜在故障,验证了所提故障预警方法的时效性。
In order to warn the potential failure of wind turbines and enhance the safety of the unit output,a wind turbine fault warning method with abnormal data reconstruction is proposed based on the supervisory control and data acquisition(SCADA)system.Firstly,the SCADA data of the wind turbines in the same wind farm are fully utilized to reconstruct the two types of target unit data of the input and output,respectively,which overcomes the problems of partial data missing and data anomaly.Secondly,a fault warning model is established using the extracted representative data,which is closer to describe the dynamic characteristics of the unit.Thirdly,the improved deterioration degree is adopted to warn the potential failures and intuitively show the phased deterioration of the unit.In the case study,the SCADA fault data of a certain wind farm were used,and the parameter settings of the proposed strategy were determined with three criteria.The results show that the proposed method can predict the potential fault of the gearbox of the wind turbine at least 3 weeks ahead,which verifies the timeliness of the proposed early warning method.
作者
刘帅
刘长良
甄成刚
Liu Shuai;Liu Changliang;Zhen Chenggang(School of Control and Computer Engineering,North China Electric Power University,Changping District,Beijing102206,China;State Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources,North China Electric Power University,Changping District,Beijing102206,China)
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2019年第8期1-11,共11页
Chinese Journal of Scientific Instrument
基金
北京市自然科学基金(4182061)
中央高校基本科研业务费专项资金(2018ZD05,9163116001)项目资助
关键词
故障预警方法
梯度提升回归树
高斯混合模型
数据分类重建
衰退指标
fault warning method
gradient boosted regression tree
Gaussian mixed model
classified data reconstruction
deterioration degree
