摘要
为提高偏航对风精度,减少由于风向波动引起偏航系统频繁操作,提出了一种基于风向样本相似度和支持向量机预测模型的风机偏航系统控制策略。针对1 MW异步双馈风力发电机组,首先对风场数据进行统计分析得到风向波动周期特性并据此建立风向样本,根据相似度原理和支持向量机算法对风向样本进行聚类分析并建立风向预测模型,根据预测结果对偏航系统进行控制。仿真结果表明,该预测模型能够预测风向变化基本趋势,该策略可大幅度提高对风精度。
A kind of fan yaw system control strategy based on fan sample similarity and support vector machine prediction model is proposed in order to improve yaw accuracy toward wind and reduce frequent operation of yaw system due to wind direction fluctuation.For 1MW Asynchronous doubly-fed wind turbine,firstly,carry out statistical analysis of wind field data and get characteristics of fluctuation cycle of wind direction and establish wind direction sample accordingly.Carry out cluster analysis according to similarity principle and algorithm of support vector machine and establish wind direction prediction model.Implement control of yaw system according to prediction results.Simulation results show the prediction model can predict basic trend of wind direction change. The strategy can improve accuracy toward wind greatly.
作者
许炳坤
XU Bingkun(China Datang Northwest Electric Power Test and Research Institute ,Xian 710000, China)
出处
《国网技术学院学报》
2017年第1期50-54,共5页
Journal of State Grid Technology College
关键词
偏航系统
相似度
数据统计
预测模型
yaw system
similarity
data statistics
prediction model
