Smart socket-based electrical appliance type inference method and device

Abstract

The invention discloses a smart socket-based electrical appliance type inference method and device, wherein the method comprises the following steps: constructing a user power data set according to electrical appliance power consumption data, and constructing an electrical appliance power consumption characteristic model library according to the electrical appliance power consumption characteristic; constructing an electrical appliance behavior data set according to the user power data set and the electrical appliance power consumption characteristic model library; constructing a characteristic vector extracting model according to the electrical appliance behavior data set so as to construct a power consumption characteristic vector data set; constructing an electrical appliance type classifier according to the power consumption characteristic vector data set and a machine learning algorithm; detecting the power consumption data of a to-be-measured electrical appliance and determiningthe electrical type of the to-be-measured electrical appliance through the model and the classifier. According to the method, the power data of the user is collected through a smart socket; by extracting the characteristics of the range of the power consumption of the household appliances, the duty ratio of the power consumption and the duration of time of power consumption, the power utilizationbehavior of the user is inferred, an electrical appliance behavior event data set is generated, and an electrical appliance type classifier is constructed based on the machine learning model, so thatthe type of the utilized electrical appliance can be inferred in real time; the robustness is high; and the method is simple and easy to implement.

Description

technical field [0001] The invention relates to the technical fields of computer, communication and information security, in particular to a method and device for inferring the type of an electrical appliance based on an intelligent socket. Background technique [0002] Smart home is a ubiquitous computing application that can provide users with context-aware automatic or assisted services in the form of ambient intelligence, remote home control, or home automation. There are three main application directions of smart home, namely comfort, health protection and security. Smart home can improve user comfort through activity recognition, automatic trigger events, remote access and control, implement health protection through local and remote monitoring, and ensure home security through user and device authentication. With the rapid popularization of smart home devices, smart homes have entered people's homes, providing people with a more convenient and comfortable life. The ...

AI Technical Summary

Problems solved by technology

[0005] However, there are two limitations in the non-intrusive power load monitoring of related technologies. One is that in order to extract various power characteristics, high-frequency power data is required

High-frequency power data increases the difficulty of data collection, increases the overhead of data storage, and processing a large amount of data increases the computational overhead and delay of the entire power analysis system

The second is that the data used by these two methods can only be obtained from the power supply company. In an environment where privacy protection is increasingly valued, this method of data acquisition is becoming more and more difficult

Images