| 引用本文: | 黎静华,黄玉金,黄乾.基于改进混沌时间序列的风电功率区间预测方法[J].电力自动化设备,2019,39(5): |
| LI Jinghua,HUANG Yujin,HUANG Qian.Interval prediction method of wind power based on improved chaotic time series[J].Electric Power Automation Equipment,2019,39(5): |
|
| 摘要: |
| 风电功率区间预测是预测给定置信水平下风电功率的上限和下限，可以反映风电功率的变化范围，为调度提供有效的辅助信息。考虑风电功率的混沌特性，提出了基于改进混沌时间序列的风电功率区间预测方法。由于风电功率具有强间歇性和波动性，传统的混沌时间序列方法在风电功率区间预测中难以获得好的聚类效果和高的预测精度，影响了功率区间预测的结果。引入蚁群聚类算法和支持向量机，利用蚁群聚类算法的强搜索能力和支持向量机的强预测能力对传统方法进行改进，获得了更好的区间预测结果。将改进方法应用于英国和德国风电场的风电功率区间预测中，对比分析改进方法与基于神经网络的功率区间预测方法和传统方法在不同置信水平下的预测结果，验证了所提改进方法的有效性。 |
| 关键词: 风电功率 区间预测 混沌时间序列 蚁群聚类算法 支持向量机 |
| DOI：10.16081/j.issn.1006-6047.2019.05.008 |
| 分类号:TM614 |
| 基金项目:国家重点研发计划支持项目(2016YFB0900100)；国家自然科学基金资助项目(51377027) |
|
| Interval prediction method of wind power based on improved chaotic time series |
|
LI Jinghua, HUANG Yujin, HUANG Qian
|
|
Guangxi Key Laboratory of Power System Optimization and Energy-saving Technology, Guangxi University, Nanning 530004, China
|
| Abstract: |
| Wind power interval prediction is to predict the upper and lower limits of wind power at a given confidence level, which can reflect the variation ranges of wind power and provide effective auxiliary information for scheduling. An interval prediction method of wind power based on improved chaotic time series is proposed considering the chao-tic characteristics of wind power. Due to the strong intermittency and fluctuation of wind power, the traditional chaotic time series method is difficult to obtain good clustering effect and high prediction accuracy in wind power interval prediction, which affects the prediction results. The ant colony clustering algorithm and support vector machine are introduced to improve the traditional method by using the strong searching ability of ant colony clustering algorithm and the strong predictive ability of support vector machine, and then better interval prediction results are obtained. The improved method is applied to the wind power interval prediction of wind farms in Britain and Germany. The interval prediction results of the improved method, the interval prediction method based on neural network and the traditional method are compared and analyzed at different confidence levels, and the validity of the proposed improved method is verified. |
| Key words: wind power interval prediction chaotic time series ant colony clustering algorithm support vector machines |
