Load characteristics discrimination analysis based on mathematical statistics and fuzzy clustering
An Luo · Electric Power · 2011
The load characteristics discrimination is an important task in load modeling.The substation characteristics clustering must be adjusted dynamically with the power grid expansion and substation growth.Two algorithms were proposed to discriminate loads by using Mahalanobis distance in mathematical statistics and fuzzy C means clustering.Based on the testing results on known characteristics of actual substation load data,the advantages and disadvantages of both algorithms are discussed.The mathematical statistics method has simple calculation and suitable when the substation number is small.The fuzzy clustering method yields more accurate results and is suitable for the case with large number of substations.The two algorithms are accurate and effective for substation load characteristics identification.