Extraction of Power Load Patterns Based on Cloud Model and Fuzzy Clustering
Song Yiyan · Power System Technology · 2014
In allusion to such defects as sensitive to initial clustering center, not convenient to determine clustering number and poor stability of the algorithm during utilizing traditional fuzzy C-Means(FCM) algorithm to extract power load patterns, starting from the morphological feature of load curve and based on cloud model and fuzzy clustering a method to extract the power load pattern is proposed. Firstly, according to the time characteristic of power load data the dimension extension for the cloud transformation is performed to make it enable to be applied in the processing of two-dimensional data with time characteristic, and the frequency distribution of typical daily load curves of power consumers is decomposed into the superposition of several normal cloud groups by multiple cloud transform and the expected vector set in all cloud models, which can mostly represent the qualitative concepts, is taken as the initial clustering center; secondly, based on the initial clustering center and the number of clustering determined by the cloud model the FCM algorithm is utilized to extract power load patterns and classify power consumers; finally, the case analysis based on actual data of a certain power grid is carried out and analysis results show that the proposed method is practicable and effective.