A non-intrusive load recognition method that integrates transient characteristics and changes between steady states

Abstract

The present invention proposes a non-intrusive load-type identification method that integrates transient characteristics and changing characteristics between steady states, prevents the problem of distinguishing the types and states of electrical appliances caused by similar steady-state characteristics, and realizes the effect of intelligent and optimized identification of electrical appliances in the system And improve the identification effect of the transient state and steady state process of the electrical appliance. The monitoring and storage method is adopted. When the amplitude of the fundamental wave current changes, the sampling period is stored. When no multiple changed sampling periods are detected, the transient process identification and steady-state process of the electrical appliance are identified through the improved CUSUM method. . Through the combination of the convolutional neural network to identify the transient process of the electrical appliances and the C-means clustering method to identify the change process between the steady states of the electrical appliances, identify the types of electrical appliances and what kind of working state they are in. The invention can construct a dynamic and personalized electrical appliance feature database for users. Finally, based on the sinusoidal superposition formula, a more accurate change of the current amplitude at each frequency between the front and rear steady states can be obtained.

Description

technical field [0001] The invention relates to the field of smart grids, in particular to a non-intrusive load identification method that integrates transient characteristics and change characteristics between steady states. Background technique [0002] With the reduction of non-renewable energy and the increase of extreme climate in the world caused by the greenhouse effect, how to improve the efficiency of energy use has become an important part of the policies of various countries to achieve environmental protection and energy conservation. As an energy source closely related to residents. In the past, residents could only understand the overall electricity consumption of the household through traditional electricity meters, and it was difficult to understand the electricity consumption of each specific appliance, so that they could choose more efficient and energy-saving appliances in a targeted manner, thereby reducing energy consumption. consumption. Therefore, it ...

AI Technical Summary

Problems solved by technology

[0006] 1. Some electrical appliances have very similar steady-state characteristics in the steady state, and it is impossible to effectively distinguish the types of electrical appliances;

[0007] 2. The dependence on the collection database is high, and it is impossible to optimize the recognition effect of electrical appliances according to changes in the household electricity environment;

[0008] 3. Due to the ideal research environment, the past research methods did not consider the internal relationship between the timing of the electrical appliance state change event and the sampling data, which would lead to a longer time due to the self-transient state of some electrical appliances. Only half of the transient events can be collected in a sampling period, resulting in providing wrong sampling samples for the recognition system, resulting in system recognition errors

Images