| 引用本文: | 马斌,张丽艳,郭成.一种变权重风电功率最优组合预测模型[J].电力系统保护与控制,2016,44(5):117-121. |
| MA Bin,ZHANG Liyan,GUO Cheng.An optimal combination forecasting model with variable weight for wind power[J].Power System Protection and Control,2016,44(5):117-121 |
|
| 摘要: |
| 针对单项预测方法的局限性，利用改进的基于灰色关联度的IOWGA算子组合预测模型，建立了一种风电功率最优组合预测模型，并通过改进多种群遗传算法(MPGA)对该模型进行优化。采用RBF神经网络法、相似日法和支持向量机(SVM)法对预测日和预测日前一日的风电功率分别进行预测，通过提出的最优组合预测模型及优化算法对预测日的24 h风电功率进行组合预测。根据云南某风电场的实测数据，进行了实例分析。结果表明，风电功率最优组合预测模型能够有效提高风电功率预测精度，具有较强的实用性。 |
| 关键词: 风电功率 最优组合预测 对数灰关联度 IOWGA算子 多种群遗传算法 |
| DOI：10.7667/PSPC150831 |
| 分类号: |
| 基金项目:云南电力试验研究院(集团)有限公司电力研究院资助项目(K-YN2014-028) |
|
| An optimal combination forecasting model with variable weight for wind power |
|
MA Bin1, ZHANG Liyan1, GUO Cheng2
|
|
1.School of Electrical Engineering, Southwest Jiaotong University, Chengdu 610031, China;2.Yunnan Electrical Power Research Institute, Kunming 650217, China
|
| Abstract: |
| Aiming at the limitations of single forecasting method, the improved combination prediction model based on IOWGA operator for degree of logarithm grey incidence is used to build an optimal combined forecasting model for wind power, and the model is optimized by improved multiple population genetic algorithm (MPGA) RBF neural network method, similar day method and support vector machine (SVM) method are respectively used to predict wind power for predicting daily and the day before, the optimal combined forecasting model is used to predict 24 h wind power for predicting daily. The actual example is analyzed, according to the measured data of a wind farm in Yunnan province, the results show that the optimal combined forecasting model could effectively improve the forecasting accuracy for wind power, and has stronger practicality. |
| Key words: wind power optimal combination forecasting degree of logarithm grey incidence IOWGA operator multiple population genetic algorithm |
