风电功率预测方法综述  

Review of Wind Power Forecasting Method

作　　者：钱政[1] 裴岩[1] 曹利宵 王婧怡[1] 荆博[2] 

机构地区：[1]北京航空航天大学仪器科学与光电工程学院,北京100191 [2]江苏大学电气信息工程学院,镇江212013

出　　处：《高电压技术》2016年第4期1047-1060,共14页High Voltage Engineering

基　　金：国家自然科学基金(61573046)~~

摘　　要：随着风电接入容量的持续增长,风力发电的间歇性和波动性对电网造成的影响越来越明显,因此风电功率预测方法的研究得到了广泛的关注。准确的风电功率预测可以给电网调度、机组组合操作、风电场运营维护等提供必要的依据。从3个方面对目前的风电功率预测方法和进展进行介绍。首先,介绍了两种确定性预测方法:仅使用历史数据的统计学习方法和使用了NWP(numerical weather prediction)数据的物理模型。其次,介绍了用于提供预测结果不确定度的概率性预测方法。最后,由于风电爬坡事件会对电网造成较大的影响,还介绍了目前风电爬坡事件预测方法的研究和进展。对现有的风电功率预测方法介绍后,提出了目前风电功率预测模型遇到的一些问题以及需要进行深入研究的方向。With the rapid development of wind power, the integration of wind power is also growing continuously. The intermittency and uncertainty of wind power may pose danger on the safety on the power system, thus the wind power needs to be forecast accurately for the safety and operation of the power system. We review current methods and advances in wind power forecasting from 3 aspects. Firstly, two kinds of deterministic forecasting methods, namely,statistical methods using only historical data and physical model using NWP（numerical weather prediction） data as input,are discussed. Then, the probabilistic forecasting methods that aim to provide uncertainty of the prediction are analyzed in detail. For the wind power ramp events might affect the safety and operation of the grid seriously, the wind power ramp event forecasting methods and advances are also discussed. Furthermore, the challenges for the application of the current wind power forecasting models and the possible approaches to address these challenges are discussed for the future research of wind power forecasting.

关 键 词：风电功率预测 统计学习方法 物理模型 确定性预测 概率性预测 爬坡事件预测 

分 类 号：TM614[电气工程—电力系统及自动化]