| 引用本文: | 王晓兰,葛鹏江.基于相似日和径向基函数神经网络的光伏阵列输出功率预测[J].电力自动化设备,2013,33(1): |
| WANG Xiaolan,GE Pengjiang.PV array output power forecasting based on similar day and RBFNN[J].Electric Power Automation Equipment,2013,33(1): |
|
| 摘要: |
| 选取太阳辐照时间、辐照强度以及气温等影响光伏阵列输出功率的主要气象因素，根据相似日的输出功率具有较强的关联度，提出选择相似日的方法，设计基于相似日和径向基函数（RBF）神经网络的光伏阵列输出功率预测模型。选取最邻近的一个相似日与待预测日气象特征向量的差值作为RBF神经网络的输入变量，神经网络的输出值即为待预测日光伏阵列输出功率。以我国西北某地光伏阵列的实测功率数据对所提模型进行训练和验证，得到预测模型的平均绝对百分误差为13.82%，均方根误差为0.405 4，验证了所提模型具有较好的精度。 |
| 关键词: 光伏阵列 输出功率 径向基函数网络 相似日 预测 模型 神经网络 |
| DOI： |
| 分类号: |
| 基金项目:国家自然科学基金资助项目（50967001）；人力资源与社会保障部留学人员科技活动项目（1003ZSB112） |
|
| PV array output power forecasting based on similar day and RBFNN |
|
WANG Xiaolan1,2, GE Pengjiang1
|
|
1.College of Electrical and Information Engineering，Lanzhou University of Technology，Lanzhou 730050，China;2.Key Laboratory of Gansu Advanced Control for Industrial Processes，Lanzhou 730050，China
|
| Abstract: |
| As the solar radiation time，solar radiation intensity and air temperature are the main influencing factors of PV array output power and there is strong correlation in output power among the similar days，a forecasting model of PV array output power is designed based on similar days and RBFNN（Radial Basis Function Neural Network），which takes the difference of meteorological feature vector between the nearest similar day and the day to be forecasted as the input variable of RBFNN and its output as the forecasted PV array output power. The proposed model is trained and verified with the measured power dada of a PV array in Northwest China. Its mean absolute percentage error obtained is 13.82 % and its root mean square error is 0.405 4，showing its high accuracy. |
| Key words: PV array output power radial basis function networks similar day forecasting models neural networks |
