| 引用本文: | 李 辉,胡姚刚,李 洋,杨 东,欧阳海黎,兰涌森,唐显虎.基于温度特征量的风电机组关键部件劣化渐变概率分析[J].电力自动化设备,2015,35(11): |
| LI Hui,HU Yaogang,LI Yang,YANG Dong,OUYANG Haili,LAN Yongsen,TANG Xianhu.Gradual deterioration probability analysis based on temperature characteristic parameters for critical components of wind turbine generator system[J].Electric Power Automation Equipment,2015,35(11): |
|
| 摘要: |
| 为了掌握风电机组关键部件劣化程度及渐进变化趋势，提出基于温度特征量的风电机组关键部件劣化渐变概率分析方法。针对采用固定阈值不能准确确定劣化度的问题，利用风电机组关键部件的温度特征量和转速信息，分别提出基于数据拟合和机群划分思路的劣化度上下限动态阈值确定方法。考虑部件劣化度会受运行工况和运行时间的影响，应用非参数核密度估计法建立风电机组关键部件劣化度的概率密度函数，提出不同监测周期内的渐变趋势概率分析方法。以实际的风电机组发电机后轴承劣化渐变情况为例，基于某实际风电场历史监测数据，对提出的动态阈值确定和部件劣化概率分析方法进行验证，并与采用固定阈值确定方法进行比较。结果表明，所提方法更能准确确定部件的劣化度，能有效分析风电机组关键部件劣化渐变趋势。 |
| 关键词: 风电机组 状态监测 动态阈值 机群划分 概率密度 劣化渐变趋势 风电 |
| DOI： |
| 分类号: |
| 基金项目:国家国际科技合作专项资助项目（2013DFG61520）；国家自然科学基金资助项目（51377184）；中央高校基本科研业务费专项基金资助项目（CDJZR12150074）；重庆市集成示范计划项目（CSTC2013JCSF70003）；重庆市研究生科研创新项目（CYB14014） |
|
| Gradual deterioration probability analysis based on temperature characteristic parameters for critical components of wind turbine generator system |
|
LI Hui1, HU Yaogang1, LI Yang1, YANG Dong1, OUYANG Haili2, LAN Yongsen2, TANG Xianhu3
|
|
1.State Key Laboratory of Power Transmission Equipment & System Security and New Technology，Chongqing University，Chongqing 400044，China;2.CSIC（Chongqing） Haizhuang Wind Power Equipment Co.，Ltd.，Chongqing 401122，China;3.Chongqing KK-QIANWEI Wind Power Equipment Co.，Ltd.，Chongqing 401121，China
|
| Abstract: |
| A method of gradual deterioration probability analysis based on the temperature characteristic parameters is proposed for the critical components of WTGS（Wind Turbine Generator System） to grasp their deterioration level and tendency. Since the fixed threshold may not be used to accurately determine the degradation degree，the concept of data fitting and turbine grouping based on the temperature characteristic parameters and rotation speeds of the critical components is proposed to set the dynamic thresholds for the upper and lower limits of degradation degree. In order to include the effects of operating condition and duration on the degradation degree，the nonparametric kernel density estimation method is applied to build the probability density function of degradation degree for the critical components and a method of gradual deterioration probability analysis is presented for different monitoring cycles. With the rear bearing of a wind-turbine generator in an actual wind farm as an example，the proposed dynamic threshold method and probability analysis method are verified based on the historical monitoring data of its gradual deterioration. Compared with the fixed threshold method，the proposed dynamic threshold method can more accurately determine the degradation degree of components and more effectively analyze the deterioration tendency of critical components. |
| Key words: wind turbines condition monitoring dynamic threshold turbine grouping probability density deterioration tendency wind power |
