变尺度时间窗口和波动特征提取的短期风电功率组合预测

发布时间：2018-08-15 13:57

【摘要】：精确的风电功率预测对保障大规模风电接入电网后电力系统的安全稳定运行具有重要意义。其中,风速的随机变化是引起风电功率波动和影响风电功率预测精度的最主要原因。针对该问题,提出一种基于变尺度时间窗口和波动特征提取的短期风电功率组合预测方法。首先,通过多重分形谱分析不同天气类型下的风速特征。然后,根据当前风速的特征量采用变尺度滑动时间窗口算法,动态地进行特征提取,由提取结果对风电历史数据进行分类,在此基础上选择特定参数建立对应的功率预测模型。为使模型在功率大幅度波动时刻的预测结果更加精确,提出了基于频谱分析的修正方法。最后,将不同天气类型下的功率预测结果与修正结果进行时序组合。算例结果表明,所述变尺度时间窗口与波动特征提取相结合的短期风电功率组合预测方法可有效提高风速波动剧烈的风电场的风电功率预测精度。
[Abstract]:Accurate wind power prediction is of great significance to ensure the safe and stable operation of power system after large-scale wind power is connected to the power network. The random variation of wind speed is the main cause of wind power fluctuation and the influence of wind power prediction accuracy. To solve this problem, a short-term wind power combination prediction method based on variable scale time window and wave feature extraction is proposed. Firstly, the wind speed characteristics of different weather types are analyzed by multifractal spectrum. Then, according to the current wind speed feature, the variable scale sliding time window algorithm is used to extract the feature dynamically, and the historical wind power data is classified by the result of the extraction. Based on this, the corresponding power prediction model is established by selecting the specific parameters. In order to make the prediction results of the model more accurate at the time of large power fluctuation, a modified method based on spectrum analysis is proposed. Finally, the power prediction results of different weather types are combined with the modified results. The numerical results show that the combined short-term wind power prediction method combined with variable scale time window and wave feature extraction can effectively improve the prediction accuracy of wind power in wind farms with severe wind speed fluctuations.
【作者单位】：中国农业大学信息与电气工程学院;国网冀北电力有限公司;中国电力科学研究院;国网冀北电力有限公司电力科学研究院;
【基金】：国家自然科学基金资助项目(51477174);国家自然科学基金中英国际合作交流基金资助项目(51711530227) 国家电网公司科技项目(5201011600TS)~~
【分类号】：TM614

【相似文献】