基于近邻传播聚类权重的黑启动方案评估方法

黑启动方案的优选和评估对电力系统在大面积停电后的快速恢复具有重要的意义。熵权法是黑启动方案评估中常用的权重确定方法,但却存在指标权重分配差别过大,无法体现决策矩阵微小变化等缺点。基于此,文中提出了一种新的权重方法——基于近邻传播聚类的权重确定方法,并将其用于黑启动方案评估。首先,利用近邻传播聚类算法对指标完全情况下的黑启动方案进行聚类,得到全指标黑启动方案聚类结果。其次,依次隐藏每个指标,再采用近邻传播聚类算法对黑启动方案进行聚类,得到缺失指标情况下的黑启动方案聚类结果。再次,根据结构相似性思想,对全指标黑启动方案聚类结果和缺失指标黑启动方案聚类结果进行相似性计算,得到指标权重,并采用线性加权法完成对所有黑启动方案的完全排序。最后,在中国广东电网黑启动数据集上对文中所提方法进行验证,实验结果表明所提方法具有较高的准确性。

Assessment Method of Black-start Scheme Based on Affinity Propagation Clustering Weight

Best selection and assessment of black-start schemes play an important role in fast restoration after a large-area blackout in power system. The entropy method is one of the most popular methods for determining the weights in black-start scheme assessment. The entropy method suffers from a range of problems such as large deviation of weight distribution, failing to reflect small changes of the decision matrix, etc. A novel weight determination method based on affinity propagation (AP) clustering is proposed, which is applied to evaluate black-start schemes. Firstly, the AP clustering algorithm is used to cluster the black-start schemes with all indices, and the clustering results based on all indices are obtained. Then, each of the indices is deleted one by one, the AP clustering algorithm is used to cluster the black-start schemes, and the clustering results based on the remaining indices are obtained. Moreover, according to the structure similarity, the similarities between the clustering results based on all indices and incomplete indices are calculated, and the weights of indices are determined. The linear weighting method is used to sort all black-start schemes. Finally,based on the data of Guangdong power system of China, experiments are carried out to validate the effectiveness of the proposed method. The results show that the proposed method has high accuracy.