基于量测的负荷建模--分类算法

负荷建模的最大困难在于负荷的时变性，该文成功地将多元统计分析中的系统聚类法引入负荷建模领域，提出了一种分类算法。该算法可分析负荷组成成分的时变规律，求解负荷组成时变特性到参数变化的映射，然后根据负荷组成的时变规律，在不同的时间采用不同的负荷模型参数，从而提高了负荷模型的精度。最后，作者给出了文中所提算法在工程中的一个应用实例，说明了该算法的有效性。

MEASUREMENT-BASED LOAD MODELING- SORTING ALGORITHM

The most difficult part in load modeling results from the time variance of loads. This paper introduces cluster analysis method, which belongs to multi-analysis and statistics to the load modeling, and presents a new sorting algorithm. The new algorithm analyzes the rule of time variance of the load components, and gives out the mapping relation from the time-varying characteristics of load components to the variance of characteristics. Then according to the time-varying rule of the load components, this algorithm models the load by different parameters at different time, which greatly improves the precision of the load modeling. Finally, the effectiveness of this new algorithm is proved by a project instance.