基于EEMD与AR建模的风电场风速预测

doi:10.3969/j.issn.1000-1158.2015.02.16

摘要
图/表
参考文献(11)
相关文章 (15)

全文: PDF (770 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要针对风速时间序列的非平稳性与非线性,提出一种基于集合经验模式分解(EEMD)与AR建模分析的风电场风速集成预测方法。首先运用EEMD对风速序列进行预处理,将其分解为一系列相对平稳的固有模态分量(IMF),突出原始风速序列的局部特征信息;然后利用AR建模对各分量进行预测分析,降低建模难度与预测成本;最后,将各分量的预测结果利用最小二乘法求得权值后进行集成得到风速序列的预测结果。风电场实测数据验证表明,相比单一的AR建模预测和基于EMD的AR集成预测,该方法有效地提高了预测的精度。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	何群
	赵文爽
	江国乾
	谢平

关键词 ：计量学, 风速, 集成预测, EEMD, 时间序列建模, 最小二乘法

Abstract：Aiming at the non-stationary and nonlinear of wind speed sequences ,an integrated method based on EEMD and AR modeling is proposed. The wind speed time series are firstly pretreated by EEMD and decomposed into a series of relatively smooth IMF components, highlighting the local characteristics of the original sequences. Then each IMF component is modeled and forecasted using AR modeling, thus reducing the difficulty of modeling and forecast costs. Eventually, the prediction results of each component are taken for integration by the least square method to get the right values. A set of wind speed data from some wind farm are verified and the results show that compared with the single AR modeling prediction and forecast based on EMD and AR integration, the proposed method can effectively improve the prediction accuracy.

Key words： Metrology Wind speed Integrated prediction EEMD Time series modeling Least square method

基金资助:河北省自然科学基金(F2011203149);
河北省高等学校科学技术研究重点项目(ZD20131080)

作者简介: 何群(1969-),男,黑龙江哈尔滨人,燕山大学副教授,研究方向为风力发电系统健康预测及故障诊断、预测控制技术及最优化算法等。hq@ysu.edu.cn

引用本文:

何群,赵文爽,江国乾,谢平. 基于EEMD与AR建模的风电场风速预测[J]. 计量学报, 2015, 36(2): 181-186.
HE Qun,ZHAO Wen-shuang,JIANG Guo-qian,XIE Ping. Wind Speed Prediction Based on EEMD Analysis and AR Modeling. Acta Metrologica Sinica, 2015, 36(2): 181-186.

链接本文:

http://jlxb.china-csm.org:81/Jwk_jlxb/CN/10.3969/j.issn.1000-1158.2015.02.16 或 http://jlxb.china-csm.org:81/Jwk_jlxb/CN/Y2015/V36/I2/181