Non-intrusive load identification method based on regularization greedy forest algorithm

Abstract

The invention relates to a non-intrusive load identification method based on a regularization greedy forest algorithm. The method specifically comprises the following steps: firstly, using residential electricity consumption data of a certain area, and preprocessing the data; secondly, selecting a V-I track as a load feature, wherein an extraction method of the track feature is that an original V-I track is converted into a two-dimensional V-I track through mapping, the correlation between a feature track graph and a label of sample data is analyzed, and the load feature related to the label of the sample data is used for improving the accuracy of load identification; and finally, carrying out load identification by using a non-intrusive load identification method based on a regularization greedy forest algorithm, and obtaining an identification result. According to the method, the load identification precision can be improved, and the model has good generalization ability and certain application value.

Description

technical field [0001] The invention relates to a non-invasive load identification method based on a regularized greedy forest algorithm, and is particularly suitable for involving in a non-invasive load monitoring system. Background technique [0002] With the widespread popularization of the smart grid, the residential user side has become one of its important consumption ends. Through the sub-item measurement and real-time feedback of the user's electricity consumption, it can guide residents to form reasonable electricity consumption habits, which plays a vital role in alleviating the energy crisis. At the same time, it helps the grid side to deeply explore the energy-saving potential and demand response potential of residents; non-intrusive load monitoring is the way to realize sub-item metering of electricity consumption, and load identification is one of the important components of non-intrusive load monitoring , which has important research significance; the present ...

AI Technical Summary

Problems solved by technology

[0002] With the widespread popularization of the smart grid, the residential user side has become one of its important consumption ends. Through the sub-item measurement and real-time feedback of the user's electricity consumption, it can guide residents to form reasonable electricity consumption habits, which plays a vital role in alleviating the energy crisis. At the same time, it helps the grid side to deeply explore the energy-saving potential and demand response potential of residents; non-intrusive load monitoring is the way to realize sub-item metering of electricity consumption, and load identification is one of the important components of non-intrusive load monitoring , which has important research significance; the present invention aims at the problem that the existing high-precision load recognition algorithm based on deep learning has high computational complexity and cannot be used for home embedded devices, and announces a non-intrusive method based on regularized greedy forest algorithm Load identification method, this method can improve the accuracy of load identification, and the model has good generalization ability, which has certain application value

Images