Non-intrusive household load real-time identification method and device based on multi-feature fusion

Abstract

The present invention relates to a multi-feature fusion-based non-intrusive type household load real-time identification method and device. The method includes the following steps that: voltage and current data at the entrance of a monitored environment are acquired in real time, basic electrical parameters are calculated; an effective load switching event is monitored according to the basic electrical parameters obtained in the first step, voltage and current sampling data acquired before and after the switching event are cached; sequence steady state and transient state features before and after the switching event are extracted according to the cached information in the previous step, the feature vector of a current event is generated; samples in a load feature library are filtered according to obtained event feature vector information, so that a candidate load sample subset is obtained; the matching degree of each of candidate load samples and the feature vector of the current event is calculated sequentially; and an optimal result is selected and is compared with a preset matching degree threshold value, and a load identification result is outputted. With the multi-feature fusion-based non-intrusive type household load real-time identification method and device of the invention adopted, real-time performance and low cost are fully considered with the accuracy of load identification ensured; and a solid foundation can be laid for follow-up practical promotion.

Description

technical field [0001] The invention relates to a non-invasive real-time household load identification method and device based on multi-feature fusion, and belongs to the technical fields of household load monitoring, energy saving and emission reduction, and the like. Background technique [0002] Household load identification technology can be divided into two major technical routes, namely intrusive and non-intrusive. Among them, the intrusive type needs to install a meter for each device, which has advantages and disadvantages such as accurate measurement data, high cost, complicated installation, and difficult maintenance. The non-intrusive type installs metering instruments at the user's power supply entrance, which is low in cost, easy to install, and suitable for online monitoring, but it is technically difficult and difficult to capture load events in real time. The load characteristic is the performance of the unique power consumption mode inside the load, that is...

AI Technical Summary

Problems solved by technology

This method needs to calculate the similarity of power transient waveforms, so it requires high-density power values ​​and a large number of matching operations, resulting in a large "storage space" and high requirements for "computing performance".

In addition, this method has poor resistance to load fluctuations and is prone to false detection

Images