Electric power load curve shape clustering method based on adaptive segmentation aggregation approximation

Abstract

The present invention discloses an electric power load curve shape clustering method based on adaptive segmentation aggregation approximation. The method comprises: dividing an original electric powerload curve data sequence into a plurality of data segments, then calculating the total number of climbing events in a fixed time window and the total number of edge points extracted based on the slope, and performing dimension reduction on the electric power load curve through the total number of climbing events and the total number of edge points; when the dimension of the electric power load curve reaches the preset dimension, selecting cluster centers in the electric power load curve according to the number of clusters, and calculating the SBD distance from the load point of the electric power load curve to each cluster center; according to the SBD distance from the load point to the cluster center, dividing the load point into the category where the nearest cluster center to the loadpoint is located, and updating the cluster center of each category; and using the updated cluster center to cluster again all the load points of the electric power load curve until the number of iterations reaches the maximum number of iterations or the load points of all categories no longer change, and outputting a clustering result.

Description

technical field [0001] The invention belongs to the field of electric power system and its automation, and in particular relates to a clustering method of electric load curve shape based on self-adaptive segmentation aggregation approximation. Background technique [0002] At present, the clustering technology based on the extraction of power consumption patterns of power users in the prior art is mainly divided into the following six categories and the combination algorithms of the following methods: [0003] (1) Based on division. Specify the number of clusters, set the initial grouping, and for a given load data set, change the initial grouping through continuous iterative relocation to achieve the optimum, such as k-means, k-modoids, etc. (2) Hierarchical clustering. Divided into agglomerative method or split method. The method of agglomeration is from bottom to top, merging similar groups in each iteration until the termination condition is met; the method of splitti...

AI Technical Summary

Problems solved by technology

[0007] Aiming at the above-mentioned deficiencies in the prior art, the power load curve shape clustering method based on adaptive segmentation aggregation approximation provided by the present invention solves the increasingly multi-dimensional data of power users, which greatly increases the computational space complexity and the difficulty of time series curve shape. The problem of performing a similarity measure

Images