基于EEMD-PSO-ELM的风电功率超短期预测
Super short-term prediction of wind power based on EEMD-PSO-ELM
毛元 1冯洋 2严岩 3陈磊 3钱勇3
作者信息
- 1. 国网南京市江北新区供电公司,江苏 南京 211800
- 2. 国网宁夏电力有限公司培训中心,宁夏 银川 750011
- 3. 国网宁夏电力有限公司电力科学研究院,宁夏 银川 750011
- 折叠
摘要
针对风电场功率不稳定特性引起风电功率预测精度不高的问题,提出 1 种基于EEMD-PSO-ELM的超短期风电功率预测方法.首先,采用集合经验模态分解(ensemble empirical mode decomposition,EEMD)将风电功率序列分解为若干个模态,从而避免了模态混叠;其次,利用相空间重构对分解得到的模态计算Hurst指数,并依据Hurst指数得到最优子序列;最后,采用粒子群算法(particle swarm optimization,PSO)-极限学习机(extreme learning machine,ELM)模型对最优子序列风电功率进行预测.以某风电场为例,采用预测模型进行分析,实验结果表明EEMD-PSO-ELM预测模型的风电功率预测精度更高.
Abstract
Addressing the problem of low wind power prediction accuracy caused by the unstable characteristics of wind farm power,a super-short-term wind power prediction method based on ensemble empirical mode decomposition(EEMD),particle swarm optimization(PSO),and extreme learning machine(ELM)is proposed.Firstly,the wind power sequence is decomposed into several modes using EEMD to avoid mode aliasing.Secondly,phase space reconstruction is used to calculate the Hurst exponent for the decomposed modes,and the optimal sub-sequence is obtained according to the Hurst exponent.Finally,the PSO-ELM model predicts the wind power for the optimal sub-sequence.Experimental results from a specific wind farm illustrate that the EEMD-PSO-ELM prediction model achieves higher accuracy in wind power forecasting.
关键词
风电场功率/集合经验模态分解/相空间重构/超短期/预测精度Key words
wind farm power/EEMD/phase space reconstruction/super short term/prediction accuracy引用本文复制引用
出版年
2024
国家科技期刊平台
NSTL
维普
万方数据