Research On Non-intrusive Power Load Monitoring And Identification Technology

With the development of power demand side management technology and energy saving consciousness,the consciousness of energy saving is gradually increasing.Non-intrusive load monitoring and identification technology is widely concerned by researchers at home and abroad.The technology has the advantages of low cost,simple communication network and easy maintenance relative to invasive load monitoring technology.At the same time,the working state and energy consumption of each load are obtained which according to comprehensive information such as energy consumption information,time sharing price and electric energy measurement.It gets effective energy saving measures,make a reasonable energy saving plan and a targeted purchase of electricity load.Therefore,using non-intrusive load monitoring and identification technology to realize the visualization of residents’ electricity load can effectively reduce the cost of electricity consumption.Energy saving and emission reduction are realized to reduce the energy crisis and environmental pollution.But most researchers mainly study the load identification technology at high frequency sampling rate.This method has high requirement for sampling equipment and large amount of analysis data.It is not conducive to the popularization of non-invasive technology.Therefore,the load identification of low frequency sampling has become an urgent problem to be solved at the present stage.In this paper,a load monitoring and identification method based on multi-feature sequence fusion under low sampling rate is proposed.A visualization demonstration platform for power consumption is set up to realize non-intrusive load monitoring power visualization based on MATLAB.The main work is as follows:(1)First of all,study the characteristics of the typical household load operation.The method of wavelet threshold denoising is used to preprocess the data.According to the sliding window algorithm,the characteristics of the power,the statistical feature and the singular value are obtained It provides a basis for load monitoring and identification to build a non intrusive load identification feature database.(2)Non-intrusive load monitoring and identification method based on multi-feature sequence fusion is proposed for non intrusive power load monitoring and identification.Firstly,the load event detection algorithm is used to judge the start and stop of the load.The 0-1 programming is used to solve the possibility of the existence of the load to reduce the calculation dimension of load identification.Then the combined power sequence and the original power sequence are obtained based on the sliding time window algorithm to extracte the characteristic of power,the statistical characteristic value of the power sequence and the singular value.The load identification is carried out by using the probabilistic neural network and the DS(Dempster-Shafer)evidence theory.The estimated value of the power consumption of each load is obtained.(3)In order to test the performance of the load identification algorithm for multi-feature sequence fusion,the results of the load identification are evaluated by accuracy rate and identification precision.The simulation experiments with single load identification,multi load identification,identification of different sampling rates and multiple user load identification are carried out.The simulation results show that the method proposed in this paper can get more than 85%of the accuracy and accuracy of the load.At the same time,the accuracy rate of this method is higher than that of the traditional algorithm.The identification of high precision of load identification at low frequency sampling frequency is realized.(4)For the visualization of non intrusive load monitoring and identification algorithms,A set of residential electricity visualization system was designed to use the GUI tool in MATLAB.The system includes modules such as human-machine interaction main interface,load event detection,load possibility combination judgment,parameter setting,load characteristic database,residential electricity real-time information and historical data query module.In addition,the multi feature sequence fusion load identification method proposed in this paper is applied.In order to guide the residents to adjust the load to use electric time and to choose energy saving electrical appliances,It is able to realize the basic function of non-invasive resident power use visualization.