相空间重构的极端学习机短期风速预测模型  

A Short-term Wind Speed Prediction Model Using Phase-space Reconstructed Extreme Learning Machine

作　　者：武峰雨[1] 乐秀璠[1] 南东亮[1] 

机构地区：[1]河海大学能源与电气学院,南京210098

出　　处：《电力系统及其自动化学报》2013年第1期136-141,共6页Proceedings of the CSU-EPSA

摘　　要：对风速进行快速、准确的预测,可以有效地减小或避免风电场对电力系统的不利影响,同时提高风电场在电力市场中的竞争能力。根据风速具有混沌特性,提出一种相空间重构的极端学习机(extreme learningmachine,ELM)的短期风速预测模型,通过确定延迟时间和嵌入维数,对样本空间进行重构,使新的样本更能反映风速变化特性,在此基础上运用ELM进行短期风速预测。与传统的预测模型相比,该方法具有学习速度快、泛化性能好等优点,为风速预测提供了新方法。A quick and accurate prediction of wind speed can effectively reduce or avoid the adverse effects of wind farms on power system, and can as well improve the competitiveness of the wind farm in the electricity market. In this paper, according to the chaotic characteristics of the wind speed, a short-term wind speed prediction model using phase-space reconstructed extreme learning machine （ELM） is put forward. The decision of the delay time and em- bedding dimension is used to reconstruct the sample space, which makes the new sample better reflect the change characteristics of wind speed. On this basis, the ELM is applied for short-term wind speed prediction. Compared with the traditional prediction model, this method has the advantages of fast learning speed and good generalization perfor- mance. Therefore, a new method is provided for wind speed prediction.

关 键 词：风力发电 短期风速预测 混沌特性 相空间重构 极端学习机 

分 类 号：TM835[电气工程—高电压与绝缘技术]