Classification of 220KV Substation Based on Daily Load

It has been well recognized that the load classification has great effects on the load model building when it applied the statistic synthesis method to construct the load model. However, it is also widely known that the load classification is a quite difficult problem due to the primal data limited and singleness, for example the daily load consumption data which obtained from SCADA are very simply. Different disposal ways for the primal data will get the different cluster results. Scarcity of checkout ways and means brings on the very difficulties of judging the cluster result which is ture and which is wrong. In this paper, two kinds. of eigenvectors abstracted from daily-load-curve are proposed. Using fuzzy cluster analysis, 90 substations with 220KV that in the middle areas of China are classified into four classes. Through the checkout method suggested by this paper, the case studies showes the efficiency.