摘要
考虑到风电功率短期预测的准确性对电网调度具有重要作用,提出了一种由改进的集成经验稳态分解(MEEMD)与基于遗传算法优化的极限学习机(GAELM)相结合的短期风功率组合预测模型,首先对原始风功率时间序列进行总体平均经验模态分解(CEEMD),通过排列熵剔除异常分量,再对剩余分量进行经验模态分解(EMD),其结果即为MEEMD分解所得分量,对分量分别建立GAELM预测模型,将各分量预测结果相加,即得到最终预测结果。对东北某风电场实测数据进行试验表明,与传统预测方法相比,组合预测模型有效提高了短期风功率预测的精确性。
In order to improve the accuracy of short-term wind power prediction,this paper proposes a hybrid shortterm wind power prediction model,which combines the modified Ensemble Empirical Mode Decomposition(MEEMD)and the Extreme Learning Machine based on genetic algorithm optimization(GAELM).Firstly,the original wind power time series is decomposed by CEEMD,the abnormal components are eliminated by Permutation Entropy.And then the residual components are decomposed by EMD.The results are the components obtained by MEEMD decomposition.The GAELM prediction model is established for the components respectively.The final prediction results are obtained by adding the prediction results of each component.The experimental results of a wind farm in Northeast China are compared with the traditional prediction method,it proves that the combined prediction method improves the accuracy of short-term wind power prediction effectively.
作者
陈籽君
丁云飞
CHEN Zi-jun;DING Yun-fei(School of Electrical Engineering.Shanghai Dianji University.Shanghai 201306,China)
出处
《水电能源科学》
北大核心
2020年第8期207-210,共4页
Water Resources and Power
关键词
短期风功率预测
GAELM
MEEMD
极限学习机
遗传算法
排列熵
short-term wind power prediction
GAELM
MEEMD
extreme learning machine
genetic algorithm
permutation entropy
