摘要
提出基于不确定相关向量机的风电功率实时预测方法,采用非参数回归方法拟合预测误差的概率密度分布,得出回归函数以及估计区间,最后由误差估计区间换算成功率估计区间。利用该文的方法对吉林省某风电场进行功率预测,研究表明,该方法能对风电场功率预测的不确定性进行有效分析。
The influence of fluctuation and randomness of wind power on the prediction of wind power is great. The real- time prediction method of wind power was put forward based on the uncertainty of the relevance vector machine. The probability density distribution of the prediction error was fitted by the nonparametric regression method. The regression function and the estimation interval were obtained. Finally, the error estimate range was transformed into the estimation interval of power. The method was used to predict the power of certain wind farm in Jilin province. The research shows that the method can effectively analyze the uncertainty of the wind power prediction.
作者
杨茂
张强
Yang Mao Zhang Qiang(School of Electronic Engineering, Northeast Dianli University, Jilin 132012, China)
出处
《太阳能学报》
EI
CAS
CSCD
北大核心
2016年第12期3143-3149,共7页
Acta Energiae Solaris Sinica
基金
国家重点基础研究发展(973)计划(2013CB228201)
国家自然科学基金(51307017)
吉林省科技发展计划(20140520129JH)
吉林省产业技术与专项开发项目(2014Y124)
关键词
风电功率
功率预测
相关向量机
不确定性估计
wind power
power prediction
relevance vector machine
uncertainty estimation
