基于灰色关联和协整理论的风功率组合预测研究  

RESEARCH ON COMBINATION WIND POWER FORECASTING BASED ON GRAY CORRELATION AND COINTEGRATION THEORY

作　　者：赵欣[1,2] 王爽心[1] 刘如九[1] 

机构地区：[1]北京交通大学机械与电子控制工程学院,北京100044 [2]唐山师范学院物理系,唐山063000

出　　处：《太阳能学报》2017年第5期1299-1306,共8页Acta Energiae Solaris Sinica

基　　金：国家自然科学基金(50776005);中央高校基本科研业务费专项(2011JBM103)

摘　　要：基于灰色关联和协整性理论,研究风电功率模型筛选与组合预测问题。分析6种单项预测方法与实际功率序列的灰关联度,并进行协整性检验,剔除冗余方法。算例分析结果表明,进行单项模型筛选能进一步提高组合预测模型的精度。然后,提出基于小世界优化的变权组合预测模型,并利用该模型对筛选后的方法进行组合预测。通过一个实例将小世界优化的变权组合预测模型与等权重平均组合预测法及协方差变权组合预测法进行仿真对比,该模型具有较高的预测精度和工作效率,可验证其在实际应用中的有效性和实用性。Based on the gray correlation and cointegration theory, the model selection and combination prediction of wind power forecasting are studied. Six kinds of prediction methods were chosen to be analyzed the gray correlation with the actual power sequence. In addition, cointegration test was adopted and the redundancy methods were removed. The numerical example results showed that the above model screening can further improve the accuracy of the model of the combination prediction. Consequently, the variable weight combination prediction model based on small world optimization which is processed by the model screening is proposed to predict the wind power. Compared with the combination prediction methods of equal weighted average and covariance variable weight based on the real data of wind power forecasting system, the proposed method is verified its effectiveness and practical utilities according to the desired accuracy and efficiency level.

关 键 词：风电功率预测 组合预测 模型筛选 变权模型 小世界优化 

分 类 号：TM743[电气工程—电力系统及自动化] TP273[自动化与计算机技术—检测技术与自动化装置]