Evaluating performance of WFA K-means and Modified Follow the leader methods for clustering load curves

Clustering is a process that partitions a set of feature vectors into clusters. There are different applications of load curves clustering in regulated and deregulated environment such as system analysis, load and price forecasting, distributed resource selection, better tariff design, etc. In this paper we evaluate performances of two clustering methods (WFA (weighted fuzzy average), K-means and modified follow the leader) for load curves classification. For evaluation and comparison we use two adequacy measures (mean index adequacy and clustering dispersion indicator) that show distinction and compactness of clusters, respectively. A novel feature of this paper is that we evaluate performances of clustering algorithms on the basis of different applications on power system.