摘要
提出基于灰色关联分析与自适应提升的天牛群优化极限学习机风电功率短期预测方法。首先,利用灰色关联分析构建训练样本集,提高历史数据与预测日时间尺度上的信息关联度。在此基础上,利用天牛群算法优化极限学习机,为极限学习机寻找最优权阈值,提高其泛化能力。最后,引入集成学习理念,通过自适应提升算法学习组合多个极限学习机弱预测器,对预测误差进行修正,实现误差权重的自分配与重组。以此构成的极限学习机强预测器可进一步提高模型的预测精度,结合西北某风电场实际数据验证该方法的有效性。
A novel method of short-term wind power prediction based on grey relational analysis and beetle swarm optimization extreme learning machine is proposed in this paper.Firstly,the gray correlation analysis is used to construct a training sample set to improve the correlation between historical data and forecasting information on the daily time scale.Furthermore,the beetle swarm optimization algorithm is used to optimize the extreme learning machine and find the optimal weight threshold for the extreme learning machine to improve its generalization ability.Finally,the concept of integrated learning is introduced,and multiple weak predictors of extreme learning machines are combined through adaptive enhancement algorithm learning to correct the prediction errors to realize the selfallocation and reorganization of error weights.The strong predictor of the extreme learning machine formed further improves the prediction accuracy of the model and the effectiveness of the method is verified by the actual data of a wind farm in Northwest China.
作者
叶家豪
魏霞
黄德启
谢丽蓉
黄晨晨
赵世成
Ye Jiahao;Wei Xia;Huang Deqi;Xie Lirong;Huang Chenchen;Zhao Shicheng(School of Electrical Engineering,Xinjiang University,Urumqi 830047,China;China Construction Eighth Engineering Bureau First Construction Co.,Ltd.,Ji’nan 250000,China)
出处
《太阳能学报》
EI
CAS
CSCD
北大核心
2022年第3期426-432,共7页
Acta Energiae Solaris Sinica
基金
国家自然科学基金(51468062)。
关键词
风电
功率预测
自适应提升
灰色关联分析
天牛群算法
极限学习机
wind power
power forecast
Adaptive Boosting
grey correlation analysis
beetle swarm algorithm
extreme learning machine