Bilateral slope DTW distance load spectrum clustering method based on LTTB dimension reduction

Abstract

The invention provides a bilateral slope DTW distance load spectrum clustering method based on LTTB dimension reduction; the method comprises the following steps: S1, collecting power load data; s2, preprocessing the power load data to obtain preprocessed data; s3, carrying out LTTB dimension reduction by utilizing the preprocessed data to obtain low-dimension data; s4, obtaining a bilateral slope distance of the low-dimension data through an angle theta between each data point and the positive and negative directions of the x axis; s5, taking the bilateral slope distance as the similarity measurement of the DTW; s6, setting a clustering number, carrying out bilateral slope DTW spectral clustering algorithm calculation on the processed load data set, obtaining a clustering label corresponding to each power load data, and outputting the clustering label. According to the method, the data storage space is reduced, the calculation time of model training is saved, and the recognition effect of clustering is improved.

Description

technical field [0001] The invention relates to a bilateral slope DTW distance load spectrum clustering algorithm based on LTTB dimension reduction. Background technique [0002] With the development of social economy and the continuous construction of smart grid, the data volume of electric load is increasing and the types are becoming more and more diverse. A large amount of power load data contains differentiated power consumption information. Power system load clustering is based on the similarity between loads, and the loads with high similarity are classified into the same category, so as to obtain different types of power usage and typical power consumption patterns, and then effectively identify different power consumption patterns and loads. characteristic. Power load clustering can be applied to various occasions such as power price division and formulation, load forecasting, load model establishment, power quality detection, etc., which is of great significance ...

AI Technical Summary

Problems solved by technology

However, the traditional PAA and PCA dimension reduction methods cannot better reflect the change characteristics of the power load curve because they are easy to lose the change information and shape characteristics of the power load data during the dimension reduction process.

[0004] Although the traditional DTW (Dynamic Time Warping) method can retain the original change characteristics of the power load data, in the process of calculating the minimum cumulative distance for similarity identification, because the power load data is high-dimensional data, the traditional DTW method does not use the reduction method. dimensional method, so the traditional DTW method has too much calculation, and the clustering effect is not good on some data sets

Images