Font Size: a A A

Electrical Operating Data Acquisition For Non-intrusive Load Monitoring

Posted on:2019-12-07Degree:MasterType:Thesis
Country:ChinaCandidate:H WuFull Text:PDF
GTID:2382330563993244Subject:Electronics and Communications Engineering
Abstract/Summary:PDF Full Text Request
Household electricity has become an important part of the society's electricity consumption in high proportion.Aiming to effectively assess the working status of electrical appliances and achieve more accurate load decomposition,non-intrusive load monitoring system(NILMS)requires to collect operating data of various electrical equipment.But unlike other fields such as speech recognition and image processing in machine learning,currently there are not so many available data sets.How to deploy the acquisition terminal in practice and acquire electrical data in multi-feature is an urgent problem to be solved.The constitution of monitoring system,data collecting method and features processing are mainly introduced.The electrical data acquisition system is the prerequisite for non-intrusive load identification.This article designed and implemented a multi-feature electrical data acquisition terminal.Firstly,the characteristics of electrical appliances in load monitoring were studied,and the selection of low-frequency active power data and high-frequency current waveform data was used as the characteristics of electrical features.Then from the overall framework of the system,the chip platform and the main hardware modules of the acquisition system are introduced.Specifically,the selection of different circuit design schemes is involved.Finally,a double-buffer realtime storage structure and a denoising method based on wavelet transform softthreshold are proposed.Those designs solve the application difficulties of data storage and processing.The electrical data acquisition system is mainly divided into three major modules: the underlying hardware,software driver layer,and data service layer.After the debugging of the hardware circuit and software optimization of the system,some experiments measured the overall aspects of the system such as power supply,doublebuffer storage,calibration precision,denoising performance based on wavelet transform.The results reveal that the data acquisition system can be put into real use,detailed usage of appliances can be obtained and it also meets the data acquisition requirements of high precision,high integrity,and high SNR.
Keywords/Search Tags:Data Acquisition, Load Monitoring, Non-intrusive, Signal Processing
PDF Full Text Request
Related items