突变风电功率识别与预测：理论、方法与应用研究

Random fluctuations of wind power bring serious challenges for stable and security operation of the large scale wind power system. Identification and prediction of mutation wind power is a universal and critical difficulty. This subject focus on the application of support vector regression model and the mutation identification theory to prediction of mutation wind power, the research contents are as follows: (1) Study on the theory and method of mutation identification, propose novel methods applied to identification of mutation wind power; (2) Study on the the theory of support vector regression model, propose novel methods which are applied to build specific kernel function and selection parameters for prediction of mutation wind power, and establish an efficient and robust model to prediction of mutation wind power; (3) Predict the fluctuation range of mutation wind power, give quantitative assessment of the risk of fluctuations on the safety of wind power grid, and provide guidance for grid scheduling and control decisions. This subject is a cross-over study on theory and methods of prediction, data mining, and security warning management of power grid, the completion of this subject has important theoretical and practical significance not only on enriching the theory of mutation identification and prediction, but also on ensuring the stable operation of the power grid.

风电功率的随机波动对大规模风电并网电力系统的安全稳定运行带来严峻挑战。其中突变风电功率的识别和预测研究是被普遍认可的关键性难题。本课题将利用突变点识别理论和支持向量回归模型，对上述问题开展研究。主要内容包括：(1) 进行突变点识别理论及方法研究，提出适用于突变风电功率识别的新方法；(2) 开展支持向量回归模型的预测理论与方法研究，提出适用于突变风电功率预测的特定核函数构造及相关参数选取的新方法，并构建高效、稳健的突变风电功率预测模型；(3) 对突变风电功率波动范围进行预测研究，定量评估其对电网安全运行的风险，为并网电网调度和控制决策提供参考。本课题是预测理论与方法、数据挖掘、电网安全预警管理的交叉研究，对丰富突变点识别与预测理论和保证并网电网的安全稳定运行具有重要的理论意义和应用价值。

风电的随机波动对大规模风电并网电力系统的安全稳定运行带来严峻挑战，提高风电的预测精度是缓解该问题的有效途径之一。本项目以提高风电的预测精度为目标，系统梳理了有关风电预测的理论和方法。在总结前人经验的基础上，通过全面分析风电的随机性和波动性的复杂变化特征，利用特征信息提取方法、数据挖掘技术和统计学习理论，构建了几种高效、稳健的风电预测模型，为风电并网电力系统的控制决策提供参考。本课题是预测理论与方法、数据挖掘、电网安全预警管理的交叉研究，对丰富风电的预测理论和保证并网电网的安全稳定运行具有重要的理论意义和应用价值。