摘要
提出基于熵权法的光伏功率组合预测模型。首先采用持续法、支持向量机法、相似数据法对光伏电站输出功率进行单一模型预测,然后使用信息熵理论的熵权法客观评价单一预测方法提供的信息量,确定各单一预测方法的组合权重,建立基于熵权法的光伏输出功率组合预测模型。通过仿真对比3种单一预测方法以及熵权法组合预测模型的预测精度。研究结果表明,基于熵权法的组合预测模型可有效识别各单一预测方法包含的信息量,确定合理的权重,提高光伏功率预测的精度。
A Combination method for PV Power forecasting Based on entropy weight method was proposed. Firstly, three different forecasting models were investigated for PV power prediction in this paper. They are persistence method, support vector machine （SVM） and prediction method based on similar data. Then entropy weight method was used to acquire the combination weights by evaluating the amount of information of the methods mentioned above objectively. Finally, the combination forecasting model of PV power was established. Case study results showed that the combination method for PV power forecasting based on entropy weight method effectively identifies the amount of information of different forecasting methods and obtains appropriate combination weights. As a result, the proposed method will improve forecasting precision.
出处
《太阳能学报》
EI
CAS
CSCD
北大核心
2014年第5期744-749,共6页
Acta Energiae Solaris Sinica
基金
国家自然科学基金(61074094
51207049)
北京市自然科学基金(4132061)
中央高校基金科研基金(12MS59
13MS25)
关键词
光伏电站
功率预测
组合预测
熵权法
PV plant
power forecasting
combined forecast
entropy weight method
