A method and system for electrical load monitoring based on 3D convolutional neural network

Abstract

The invention discloses an electrical load monitoring method based on a 3D convolutional neural network, which includes: collecting voltage data and current data of electrical equipment in a steady state, and performing filtering and sampling processing on the voltage and current data; The processed voltage data and current data of the electrical equipment generate voltage-current trajectory curves of the electrical equipment at different times, and normalize and image binarize the voltage-current trajectory curves at each time. The binary images at different moments in one cycle of the load are superimposed to construct the input layer. Based on the neural network structure of 3D convolution, the time and depth parameters representing the convolution and pooling operations are introduced; the present invention is based on the 3D convolutional neural network, and combines the spatial features of continuous frames by sliding the 3D convolution kernel on the time axis to mine continuous The spatio-temporal characteristics of the sub-images can effectively utilize the spatial shape characteristics of the voltage and current traces and the information in the circulation direction, thereby improving the monitoring accuracy.

Description

technical field [0001] The invention belongs to the technical field of deep learning and pattern recognition, and more specifically relates to an electrical load monitoring method and system based on a 3D convolutional neural network. Background technique [0002] Nowadays, efficient use of power resources is the key to solving the problem of excessive energy consumption, and an important means in the process of effective use of power resources is to monitor electrical loads (such as household appliances, etc.) in real time to obtain power consumption information, which not only has It helps the power supplier to analyze the user's power consumption behavior through the power consumption information to allocate power energy, or to prevent fires caused by irregular power consumption according to the power consumption equipment used by the user, and the user can also use electricity Information to improve their own electricity consumption. [0003] The existing electrical loa...

AI Technical Summary

Problems solved by technology

[0005] In view of the above defects or improvement needs of the prior art, the present invention provides a 3D convolutional neural network-based electrical load monitoring method and system, the purpose of which is to solve the problems existing in the existing electrical load monitoring method based on transient characteristics The technical problems of high system energy consumption and high detection cost, as well as the technical problems of low recognition rate, poor model robustness and practicability in existing electrical load monitoring methods based on machine learning algorithms

Images