Open AccessResearch on Wind Deviation Detection Based on DENCLUE Abnormal Working Condition Filtering
Author(s) -
Xianlong Yin,
Yuanming Shi,
Xinda Xu,
Yongchun Duan,
Yaowu Jia,
Gang Chen,
Xinyue Zhang,
Feifei Yin
Publication year - 2020
Publication title -
iop conference series. earth and environmental science
Language(s) - English
Resource type - Journals
eISSN - 1755-1307
pISSN - 1755-1315
DOI - 10.1088/1755-1315/617/1/012015
Subject(s) - wind speed , smoothing , wind power , control theory (sociology) , cluster analysis , standard deviation , computer science , scada , mathematics , statistics , meteorology , engineering , artificial intelligence , physics , control (management) , electrical engineering
Aiming at the problem that the wind vane of wind turbines has a deviation to the wind, which damages the power generation efficiency of the unit, a filtering method of abnormal working conditions based on DENCLUE density clustering is proposed, which mainly included two stages of working condition screening and calculation of the angle of the wind deviation. Firstly, the density clustering algorithm based on DENCLUE is used to filter the working conditions of the data and to filter out the working condition data of abnormal power generation. Then, the data is further processed, including data smoothing and wind speed binning. Furthermore, according to the relationship between the output power of the unit and the yaw angle of the wind deviation, a regression model is established to obtain the wind deviation angle of the unit. Finally, the method is verified through the actual operation of SCADA data. The results show that after filtering the working conditions, more stable output can be obtained and the performance of the unit can be obviously improved.