基于特征分量组合的多小水电地区日负荷预测  

Daily Load Prediction with Abundant Small Hydropower Based on Characteristic Components Combination

作　　者：肖胜贤 伍永刚[1] 章国勇[1] 胡斌奇[2] 成涛[2] 

机构地区：[1]华中科技大学水电与数字化工程学院,武汉430074 [2]湖南省电力公司电力调度通信局,长沙410007

出　　处：《中国农村水利水电》2015年第7期161-165,共5页China Rural Water and Hydropower

基　　金：国家自然科学基金(1379081);国家电网公司科技项目(DKJS-13-00220)

摘　　要：随着小水电大规模并网,其无序发电对电网造成的冲击越加不可忽视,提高小水电负荷预测的准确率对掌握小水电的发电情况及电网调度有着重要意义。将小水电预测日的负荷解耦为基值和标幺值两部分进行预测,并引入云模型方法对基值的偏差进行拟合。同时,针对标幺化后的负荷曲线,采用集成经验模态获取其特征分量,并通过对相似日和日前负荷曲线的特征向量进行交叉组合以快速获得预测日的标幺曲线。通过对湖南某富小水电地区进行实例分析,验证了本预测方法的有效性。With the large-scale integration of small hydropower stations ,the grid will be greatly impacted due to its uncertainties .In tackling this issue ,predicting small hydropower generation load accurately and reliably becomes a meaningful tool to the power sys‐tem dispatching .In this paper ,a new load forecasting method based on the characteristic components combination of small hydro‐power loads is proposed .Based on this method ,the load curve is decoupled into maximum load and per unit value .For the prediction of maximum load ,the cloud model is employed to analyze prediction errors .For the prediction of per unit value ,the ensemble empir‐ical model decomposition （EEMD） is applied for the similar day and day-ahead load curves to obtain the characteristic components , which are combined to form the prediction values .The effectiveness of proposed method is finally demonstrated by being applied to a real region with abundant small hydropower in Hunan Province .

关 键 词：小水电 负荷预测 云模型 集成经验模态分解 

分 类 号：TV742[水利工程—水利水电工程]