The Application of Temporal Pattern Clustering Algorithms in DSM

With the development of China electric power uniform reform, DSM (demand side management) became an important tool to solve the energy efficiency problem. The important aid of DSM is load management. The fluctuation analysis of load curve is the foundation of load management. In this paper, a new temporal pattern clustering algorithm of power load curve is proposed. Comparing with traditional algorithms, this algorithm didn´t compare the concrete data of power load of same time point between different customers, but compare the curve shape or the fluctuation trend to classify the whole customer load curve. By using these methods, we can divide all the customer´s daily power load into some class. The test show the correct and efficiency of this algorithms