The invention provides a multi-feature fusion-based power load identification method and system The method comprises the steps of monitoring an active power at a power utilization inlet of a user; solving the active power by integer programming to determine a probably-existing load at the moment; acquiring a power sequence by a sliding window algorithm on the active power; extracting statistical characteristic and singular value characteristic according to the power sequence; solving similarity with each characteristic of the probably-existing load by Euclidean distance according to the statistical characteristic and the singular value characteristic; converting multi-similarity of the similarities of the statistical characteristic and the singular value characteristic to comprehensive similarity by a linear weighting sum method; and comparing the maximum value of the comprehensive similarity with a set threshold value to determine the load running at the moment With the adoption of integer programming, the calculation dimensionality is reduced; and by multi-feature fusion, the load at the power utilization inlet of the user can be accurately identified, and the method has robustness on fluctuation of the load power

Multi-feature fusion-based power load identification method and system

The invention relates to a load switch testing device based on power. The load switch testing device comprises the components of a current sensor which is used for testing a current analog signal in household power consumption and transmits the current analog signal to an A/D conversion module; a voltage sensor which is used for testing a voltage signal in household power consumption and transmits the voltage signal to the A/D conversion module; the A/D conversion module which is used for converting the current analog signal and a voltage analog signal to a current digital signal and a voltage digital signal, and transmitting the signals to an ARM embedded core processing module; the ARM embedded core processing module which is used for processing the current digital signal and the voltage digital signal, acquiring a power strength change value between two adjacent periods, comparing the power strength change value with a preset threshold, determining a load switching event and extracting signal characteristic data of the current digital signal and the voltage digital signal; and a memory module which is used for storing the current digital signal, the voltage digital signal and the signal characteristic data.

Load switch testing method and load switch testing method based on power

The invention discloses a self-adaptive threshold load turning-on and turning-off detection and identification method and system. The method includes the following steps: collecting current analog signals and voltage analog signals of power utilization of user households; processing the current analog signals and the voltage analog signals, and obtaining current digital signals and voltage digital signals; calculating a power intensity difference value between two adjacent periods on the basis of the current digital signals and the voltage digital signals, and classifying loads according to the power difference value; calculating a current intensity difference value of the two adjacent periods, comparing the current intensity difference value with a preset threshold, and determining whether turning-on and turning-off events of the loads occur; extracting signals features on the basis of the voltage digital signals and the current digital signals after the turning-on and turning-off events of the loads occur; and classifying and identifying the loads according to the signal features, and outputting types of the loads. Classification is performed according to different types of load power, different comparison thresholds are set, and thus the load turning-on and turning-off event detection accuracy is improved.

Self-adaptive threshold load turning-on and turning-off detection and identification method and system

The invention relates to a non-invasive type microwave oven starting identification method based on active two-stage uplift. The method comprises the following steps: 1, according to a set sampling frequency, sampling voltages and currents of a total power supply inlet wire of an indoor power grid; 2, setting a time window, i.e., calculating an average active power value corresponding to the time window at the total power supply inlet wire, wherein the average active power value is successively an element forming an average active power sequence; 3, according to the average active power sequence, calculating an active power differential sequence composed of an active power differential value; and 4, according to the active power differential sequence, determining a two-stage active power uplift occurs, then according to time features of two-state active power upright increment features, determining whether a microwave oven is started in the indoor power grid, and if no electromagnetic ovens are started, returning to the third step. According to the invention, active power regular uplift features when the microwave oven is started are taken into full consideration, and the accuracy of microwave oven non-invasive identification is improved.

Non-invasive type microwave oven starting identification method based on active two-stage uplift

The invention discloses a distribution network monitoring energy saving system, and belongs to the technical field of distribution network monitoring energy saving The system can reduce the energy consumption of a distribution network and can improve the electric energy quality of users A database platform is in communication connection with an electric energy meter, the electric energy meter is connected with a non-intrusive load monitoring device, the non-intrusive load monitoring device is connected with a user capacitor, a reactive power compensation end of a reactive power compensation device is connected on a line between the non-intrusive load monitoring device and the user capacitor, and a control end of the reactive power compensation device is in communication connection with a network control platform; a non-intrusive load monitoring decomposition module is connected with the database platform, a user behavior mode analysis and load prediction module is connected with the non-intrusive load monitoring decomposition module, a power grid weak point judgment and remote equipment control module is connected with the user behavior mode analysis and load prediction module, and the network control platform is connected with the power grid weak point judgment and remote equipment control module

Distribution network monitoring energy saving system

The invention discloses a load system for a household intelligent electric meter based on a similarity algorithm, which comprises the household intelligent electric meter, a steady-state power detection module, a transient-state power detection module, a comparator, a central processing unit, a communication module, a database and a server The invention also discloses a load identification method for the load identification system for the household intelligent electric meter based on the similarity algorithm The method is technically characterized in that steady-state and transient-state power features in the case of normal work of the household appliance are taken as feature parameters; through comparing and measuring the similarity between the feature parameters and a household appliance load feature parameter database, the type and the working state of the power load are determined; and then, through matching switching events, the corresponding power consumption of each household appliance is obtained Compared with the prior art, the purpose of identifying typical resistive and non-resistive appliances in the home in no need of a sample data training or learning process in advance is realized, a power list like a telephone list is provided for the user, and the user can adjust and plan use of the household appliances for energy conservation

Chen Zong

Papers

Load identification system for household intelligent electric meter based on similarity algorithm and load identification method thereof