基于粒子群优化的核极限学习机模型的风电功率区间预测方法.pdf

第 35 卷 增刊 中 国 电 机 工 程 学 报 Vol.35 Supplement Sep.30, 2015 146 2015 年 9 月 30 日 Proceedings of the CSEE ©2015 Chin.Soc.for Elec.Eng. DOI ：10.13334/j.0258-8013.pcsee.2015.S.020 文章编号：0258-8013 (2015) S-0146-08 中图分类号：TM 614 基于粒子群优化的核极限学习机模型的 风电功率区间预测方法 杨锡运，关文渊，刘玉奇，肖运启 (华北电力大学) Prediction Intervals Forecasts of Wind Power based on PSO-KELM YANG Xiyun, GUAN Wenyuan, LIU Yuqi, XIAO Yunqi (North China Electric Power University) ABSTRACT: Wind power forecasting is essential to the 基于粒子群优化的核极限学习机(PSO-KELM)模型，用于风 management of power system. However, traditional point 电功率区间预测。通过核极限学习机(KELM)建立预测模型， forecasting of wind power can hardly accurate due to the 采用粒子群算法对 KELM 的输出权值进行优化，寻找最优 stochastic nature of wind. Providing prediction intervals (PIs) 预测区间上下限，充分利用了 KELM 学习速度快、泛化能 by giving a range of potential outcomes for a given probability 力强的优点，实现了对风电功率的快速区间预测。通过与 was important to assess the potential uncertainties for such PSO-ELM 模型对比分析风电场在不同置信水平下的概率预 forecasting result. In this paper, to achieve better short-term 测结果，发现 PSO-KELM 模型的预测精度更高，速度更快， wind power PIs, a probabilistic PI model was built based on 能够为风电功率区间预测及风电并网安全稳定运行提供决 particle swarm optimization kernel extreme learning machine 策支持。 (PSO-KELM). It was established through intelligent algorithm 关键词：风电功率；区间预测；核极限学习机；粒子群 KELM so as to finding the best upper and lower limit of optimal PIs, which optimized the output weights of KELM by 0 引言 PSO algorithm. Finally realized directly forecasting prediction