Power load curve analysis method based on kmeans algorithm

Abstract

The invention discloses a power load curve analysis method based on a kmeans algorithm. The method comprises the following steps: selecting load data of a certain year as an analysis sample; taking each month as a standard unit, calculating five indexes including the maximum value, the minimum value, the maximum value occurrence time, the minimum value occurrence time and the peak-valley difference rate of monthly load characteristic samples of the sample data, performing normalization processing, and finally performing clustering by using a kmeans clustering method; the determination criterion of the class number of each group is the classification number which can be determined according to the fact that the distance between 90% of samples in the cluster and the class center is smaller than a certain specific error value, the samples which are far away from the sample data of the class center and are not isolated points are selected as typical daily load characteristics, and a classified power load curve is obtained. According to the electric power measurement load curve analysis method based on the kmeans clustering algorithm, when the kmeans clustering analysis method is used for selecting load characteristics, the clustering effect is related to whether data have obvious classification characteristics or not, and when sample data have the obvious classification characteristics, the clustering effect is related to the data when the sample data have the obvious classification characteristics.

Description

technical field [0001] The invention relates to the technical field of electric power engineering, in particular to a method for analyzing electric power load curve based on kmeans algorithm. Background technique [0002] Power load curve identification is the basis for the power dispatching control center to achieve orderly power dispatching, and it is also the basic content necessary for power market-oriented operations; accurate curve identification and timely detection of load mutation points can help dispatchers prepare for dispatching in a timely manner and improve Energy utilization efficiency and power supply reliability; under the guarantee of normal production and living conditions in the society, the cost of power generation can be effectively reduced, and the economic and social benefits can be improved. The traditional load characteristics selection method has a large error. Due to the different load characteristics in different regions, different regions The lo...

AI Technical Summary

Problems solved by technology

[0002] Power load curve identification is the basis for the power dispatching control center to realize orderly power consumption dispatching, and it is also the basic content necessary for power market-oriented operation; accurate curve identification and timely discovery of load mutation points can help dispatchers prepare for dispatching in time and improve Energy utilization efficiency and power supply reliability; under the condition of ensuring the normal production and living conditions of the society, effectively reduce the cost of power generation, improve economic and social benefits, the problem of large errors in the selection of traditional load characteristics, due to the different load characteristics of different regions, different regions The load is affected by external factors such as the atmosphere to different degrees. If the influence of external factors on the load characteristics is not considered, and the user’s load characteristics are directly analyzed, the load characteristic analysis results will deviate greatly from the actual situation.

Images