电力负荷动特性分类方法研究

A Classification Method for Aggregated Load Dynamic Characteristics

负荷特性的分类与综合是负荷模型实用化的关键,是解决负荷建模中负荷特性随机时变性问题和区域分散性问题的有效途径。文中深入分析了负荷动特性分类方法,系统阐述了负荷动特性分类特征向量的种类和基本要求。明确指出:以实测响应空间作为特征向量空间必须对实测响应进行时间坐标的标准化;以模型参数空间作为特征向量空间必须选择参数稳定性好的辨识算法;以标准电压激励下的模型响应空间作为特征向量空间必须选择具有好的内插外推能力的模型结构。最后,通过一个含有20个现场实测样本的分类实例比较和分析了3种特征向量空间的分类结果,并推荐使用实测响应空间作为负荷动特性分类的特征向量空间。

The classification and synthesis of power aggregated load characteristics is the linchpin of load model utilization and an effective way of resolving the random time variance problem and area decentralization problem in aggregated load modeling. In this paper, an in-depth analysis is made on the load dynamic characteristics classifying methods and several kinds of characteristic vector space for describing aggregated load characteristics are systematically discussed. It is pointed out that all the measured response samples must be standardized in time coordinates when the measured response space is used as the characteristic vector space. The parameter identification algorithm should have high parameter stability when the model parameter space is selected as the characteristic vector space and the model structure has to possess excellent interpolation and extrapolation ability when the model response space gained by a standard voltage excitation is used as the characteristic vector space. Finally, based on a classification project instance for aggregated load dynamic characteristics including 20 filed measured data samples, the field measured response space is recommended as the characteristic vector space to classify the dynamic load characteristics.