CN108318770A - A kind of electric appliance sorting technique - Google Patents

A kind of electric appliance sorting technique Download PDF

Info

Publication number
CN108318770A
CN108318770A CN201810455735.1A CN201810455735A CN108318770A CN 108318770 A CN108318770 A CN 108318770A CN 201810455735 A CN201810455735 A CN 201810455735A CN 108318770 A CN108318770 A CN 108318770A
Authority
CN
China
Prior art keywords
electric appliance
current
appliance
load
classifier
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201810455735.1A
Other languages
Chinese (zh)
Inventor
凌云
肖伸平
郭艳杰
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hunan University of Technology
Original Assignee
Hunan University of Technology
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Hunan University of Technology filed Critical Hunan University of Technology
Priority to CN201810455735.1A priority Critical patent/CN108318770A/en
Priority to CN201610213373.6A priority patent/CN105866581B/en
Publication of CN108318770A publication Critical patent/CN108318770A/en
Pending legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01RMEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01RMEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R19/00Arrangements for measuring currents or voltages or for indicating presence or sign thereof
    • G01R19/25Arrangements for measuring currents or voltages or for indicating presence or sign thereof using digital measurement techniques
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06KRECOGNITION OF DATA; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
    • G06K9/00Methods or arrangements for reading or recognising printed or written characters or for recognising patterns, e.g. fingerprints
    • G06K9/62Methods or arrangements for recognition using electronic means
    • G06K9/6267Classification techniques
    • G06K9/6268Classification techniques relating to the classification paradigm, e.g. parametric or non-parametric approaches
    • G06K9/6277Classification techniques relating to the classification paradigm, e.g. parametric or non-parametric approaches based on a parametric (probabilistic) model, e.g. based on Neyman-Pearson lemma, likelihood ratio, Receiver Operating Characteristic [ROC] curve plotting a False Acceptance Rate [FAR] versus a False Reject Rate [FRR]
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06KRECOGNITION OF DATA; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
    • G06K9/00Methods or arrangements for reading or recognising printed or written characters or for recognising patterns, e.g. fingerprints
    • G06K9/62Methods or arrangements for recognition using electronic means
    • G06K9/6267Classification techniques
    • G06K9/6268Classification techniques relating to the classification paradigm, e.g. parametric or non-parametric approaches
    • G06K9/6277Classification techniques relating to the classification paradigm, e.g. parametric or non-parametric approaches based on a parametric (probabilistic) model, e.g. based on Neyman-Pearson lemma, likelihood ratio, Receiver Operating Characteristic [ROC] curve plotting a False Acceptance Rate [FAR] versus a False Reject Rate [FRR]
    • G06K9/6278Bayesian classification
    • GPHYSICS
    • G08SIGNALLING
    • G08CTRANSMISSION SYSTEMS FOR MEASURED VALUES, CONTROL OR SIMILAR SIGNALS
    • G08C17/00Arrangements for transmitting signals characterised by the use of a wireless electrical link
    • G08C17/02Arrangements for transmitting signals characterised by the use of a wireless electrical link using a radio link

Abstract

A kind of electric appliance sorting technique, the characteristics of classification is identified, decision tree classifier and Bayes classifier are taken into account using the assembled classifier including decision tree classifier and Bayes classifier, carry out comprehensive identification, and recognition accuracy is high;Identification feature of the fundamental voltage current and phase difference and load current spectrum signature of the starting current feature, electric appliance that use electric appliance simultaneously as appliance type, characteristic information are abundant;Fundamental voltage current and phase difference and starting current feature, the load current spectrum signature acquisition methods of offer are simple, reliable.The electric appliance sorting technique can be used in some collective public places for needing to carry out electrical appliance management such as collective of students dormitory, large-scale country fair, can be used for other occasions for needing to carry out electrical equipment management for needing to carry out appliance type identification with statistics.

Description

A kind of electric appliance sorting technique
Present patent application is divisional application, and application No. is 201610213373.6, the applying date is in April, 2016 for original bill 8 days, a kind of entitled appliance type recognition methods.
Technical field
The present invention relates to a kind of electrical equipment identification sorter and methods, more particularly, to a kind of electric appliance sorting technique.
Background technology
Currently, the electric appliance load property identification method of mainstream includes the electric appliance load based on bearing power coefficient of colligation algorithm Recognition methods, the electric appliance load recognition methods based on electromagnetic induction, the electric appliance load recognition methods based on neural network algorithm, base Become the electric appliance load recognition methods etc. of scaling method in cyclic dispersion.Various methods can be to realize that electric appliance is negative to a certain degree The identification of property is carried, but since characteristic properties are single, means of identification is single, generally existing generalization ability is inadequate and cannot be completely accurate Really the problem of identification.
Invention content
It is an object of the present invention to for the defect of present prior art, a kind of electricity that can realize efficient identification is provided Device sorting technique.The electric appliance sorting technique is using the assembled classifier progress for including decision tree classifier and Bayes classifier Appliance type identifies;The input feature vector of the assembled classifier include the starting current feature of electric appliance, electric appliance load current frequency The fundamental voltage current and phase difference of spectrum signature and electric appliance.
In the assembled classifier, decision tree classifier is main grader, and Bayes classifier is auxiliary grader.It is described Assembled classifier, which carries out appliance type, which knows method for distinguishing, is:When Main classification device successfully realizes appliance type identification, Main classification device Appliance type recognition result be assembled classifier recognition result;When Main classification device fails to realize appliance type identification, and master The recognition result of grader is 2 kinds of 2 kinds or two or more electric type that either two or more appliance type exports Main classification device In type recognition result, the highest appliance type of probability identifies knot as the appliance type of assembled classifier in the output of subsidiary classification device Fruit;It is identified when Main classification device fails realization appliance type, and fails to provide the electric type of identification in the recognition result of Main classification device When type, the highest appliance type of probability is as the appliance type recognition result of assembled classifier during subsidiary classification device is exported.
The load current spectrum signature is prepared by the following:
Step 1: obtaining the steady state current signals of electric appliance load, and it is converted into corresponding steady-state current digital signal;
Step 2: carrying out Fourier transform to steady-state current digital signal, load current spectral characteristic is obtained;
Step 3: using the odd harmonic signal relative magnitude that overtone order in load current spectral characteristic is n times as negative Carry current spectrum feature, wherein n=1,3 ..., M;The M indicates harmonic wave highest number and M is more than or equal to 3.The harmonic wave letter Number relative magnitude is the ratio of harmonic signal amplitude and electric appliance load steady-state current virtual value.
The fundamental voltage current and phase difference of the electric appliance is prepared by the following:
Step 1., after electric appliance load enters stable state, the synchronous steady state voltage signal for obtaining electric appliance load, stable state electricity Signal is flowed, and is converted into corresponding steady state voltage digital signal, steady-state current digital signal;
2., to steady state voltage digital signal, steady-state current digital signal step carries out digital filtering respectively, extract fundamental wave Voltage signal, fundamental current signal;
Step 3., analysis calculate the phase difference between fundamental voltage signal and fundamental current signal, by fundamental voltage signal Fundamental voltage current and phase difference of the phase difference as electric appliance between fundamental current signal.
The judgement after electric appliance load enters stable state, according to each power frequency within nearest N number of power frequency period The load current virtual value in period carries out, and specific method is:Continuous sampling is carried out to the load current of electric appliance, with power frequency period For unit computational load current effective value and preserve;Calculate the average value of the load current virtual value of N number of power frequency period recently;When The load current virtual value and the load current of N number of power frequency period of each power frequency period within nearest N number of power frequency period have The average value of valid value compares, and when fluctuating range is respectively less than the relative error range E set, judgement electric appliance load enters stable shape State.The value range of the E is 2%-20%, and the value range of N is 50-500.
The starting current feature includes start-up course time, starting current maximum value, starting current maximum value time, is led to Cross following methods acquisition:
Before step 1, appliance starting, starts the load current continuous sampling to electric appliance and load current size is sentenced It is disconnected;When load current virtual value is more than ε, judgement electric appliance starts to start and turns to step 2;The ε is the numerical value more than 0;
Step 2, after electric appliance load enters stable state, turn to step 3;
It is step 3, electric using the average value of the load current virtual value within nearest N number of power frequency period as electric appliance load stable state Flow virtual value;Electric appliance is started into Startup time to the time between nearest N number of power frequency period initial time as when start-up course Between;Electric appliance is started into the time within Startup time to start-up course time between the maximum power frequency period of load current virtual value As the starting current maximum value time;By the load current virtual value and electric appliance of power frequency period where the starting current maximum value time Ratio between load steady state current effective value is as starting current maximum value.
The input feature vector of the assembled classifier further includes electric appliance load steady-state current virtual value.
The electric appliance sorting technique is known by the appliance type including information acquisition module, message processing module, communication module Other device is realized.Described information acquisition module is used to acquire the load current of electric appliance and load voltage information and send to information Manage module;Described information processing module carries out appliance type identification according to the information of input;The communication module is believed for sending The appliance type recognition result of processing module is ceased to host computer.
Described information acquisition module includes current sensor, preamplifier, filter, A/D converter;At described information The core for managing module is DSP, is either that ARM is either microcontroller or is FPGA.
The A/D converter that the core of message processing module includes may be used in the A/D converter.
Described information acquisition module, message processing module, communication module all or part of function be integrated in a piece of SoC In (System on Chip, system on chip).
The communication module also receives the related work instruction of host computer;Communication between the communication module and host computer Mode includes communication and wired communication mode;The communication include ZigBee, bluetooth, WiFi, 433MHz numbers pass mode;The wired communication mode includes 485 buses, CAN bus, internet, power carrier mode.
The beneficial effects of the invention are as follows:It is carried out using the assembled classifier including decision tree classifier and Bayes classifier The characteristics of identifying classification, taking into account decision tree classifier and Bayes classifier carries out comprehensive identification, and generalization ability and identification are accurately Rate is high;Simultaneously using starting current feature, the load current spectrum signature of electric appliance and the fundamental voltage electric current of electric appliance of electric appliance Identification feature of the phase difference as the appliance type identification device, characteristic information are abundant;There is provided include the start-up course time, Starting current maximum value, the starting current characteristic-acquisition method including the starting current maximum value time and load current frequency spectrum Characteristic-acquisition method is simple, reliable.
Description of the drawings
Fig. 1 is the device embodiment structural schematic diagram for realizing electric appliance sorting technique;
Fig. 2 is the start-up course current waveform of incandescent lamp desk lamp;
Fig. 3 is the start-up course current waveform of the resistive loads such as resistance furnace;
Fig. 4 is the start-up course current waveform of monophase machine class load;
Fig. 5 is computer and the start-up course current waveform that Switching Power Supply class loads;
Fig. 6 is electric appliance sorting technique flow chart.
Specific implementation mode
Below in conjunction with attached drawing, the invention will be further described.
Fig. 1 is to realize the device embodiment structural schematic diagram of electric appliance sorting technique, including information acquisition module 101, information Processing module 102, communication module 103.
Information acquisition module 102 is used to acquire the load voltage of electric appliance, load current and turns load voltage, load current Change voltage digital signal, current digital signal into, voltage digital signal, current digital signal are sent to message processing module 102. Information acquisition module includes the compositions such as voltage sensor, current sensor, preamplifier, filter, A/D converter portion Point, it is respectively completed sensing, amplification, filtering and the analog-digital conversion function of load current signal.When load voltage, load current range When larger, the preamplifier with programmable function can be selected, or is further added by an independent journey before A/D converter Amplifier is controlled, the load current larger to range carries out Discrete control amplification, make the voltage signal range for being input to A/D converter It is maintained at rational section, ensures conversion accuracy.Filter avoids spectral aliasing for filtering out high fdrequency component.
Voltage digital signal, current digital signal of the message processing module 102 according to input, using including decision tree classification The assembled classifier of device and Bayes classifier realizes appliance type identification.The input feature vector of assembled classifier includes opening for electric appliance The fundamental voltage current and phase difference of the load current spectrum signature and electric appliance of streaming current feature and electric appliance.Message processing module 102 Core be DSP, ARM, microcontroller, or be FPGA.When including A/D converter and the A/ in the core of message processing module When D converters are met the requirements, the A/D converter in information acquisition module 101 may be used in the core of message processing module 102 Including A/D converter.
Communication module 103 is sent to host computer for realizing the communication between host computer, by recognition result.Communication module Communication mode between 102 and host computer includes communication and wired communication mode, the side wireless communication that may be used Formula includes the modes such as ZigBee, bluetooth, WiFi, 433MHz number biography, and the wired communication mode that may be used includes 485 buses, CAN The modes such as bus, internet, power carrier.Communication module 103 can also receive the related work instruction of host computer, complete specified Task.Host computer can be the server of administrative department, can also be that various work stations or various movements are whole End.
Information acquisition module 101, message processing module 102, communication module 103 all or part of function can integrate On a piece of SoC, reduces device volume, facilitate installation.
Different electrical equipments has different starting current features.It is illustrated in figure 2 the start-up course of incandescent lamp desk lamp Current waveform.Incandescent lamp is that filament electrified regulation to incandescent state is sent out the electric light source of visible light using heat radiation.Incandescent lamp Filament usually manufactured with heat safe tungsten, but the resistance of tungsten varies with temperature greatly, with RtIndicate tungsten filament at t DEG C Resistance, with R0Indicate resistance of the tungsten filament at 0 DEG C, then the two has following relationships
Rt=R0(1+0.0045t)
For example, set the temperature of the filament (tungsten filament) of incandescent lamp in normal work as 2000 DEG C, one " 220V 100W's " The resistance when filament of incandescent lamp is worked normally at 2000 DEG C is
Its 0 DEG C resistance in no power is
Its 20 DEG C resistance in no power is
R20=R0(1+0.0045t)=52.8 Ω
I.e. incandescent lamp is in 9 times of the immediate current for starting energization more than its rated current, and maximum starting current is happened at Startup time.With the raising of incandescent lamp tungsten filament temperature, the load current of incandescent lamp exponentially reduces, subsequently into steady Determine state.
If electric appliance load steady-state current virtual value is IW, and define electric appliance load current effective value and enter electric appliance load stable state Within the relative error range of one setting of current effective value and stablize within this relative error range, then electric appliance load Into stable state.Relative error range can be set as 10%, can also be set as the 2%- such as 2%, 5%, 15%, 20% Value between 20%.In Fig. 2, the relative error range that sets is 10%, when the load current of incandescent lamp exponentially subtracts It is small to arrive its IW10% error range when, T at the time of as in Fig. 2S, start-up course terminates.The start-up course time of incandescent lamp is TS。IWFor virtual value.
Select the startup electricity of start-up course time, starting current maximum value I*, starting current maximum value time as electric appliance Flow feature;Starting current maximum value is per unit value, i.e. starting current maximum value I* is the maximum virtual value I of starting currentMWith electricity Device load steady state current effective value IWRatio.
In Fig. 2, the start-up course time of incandescent lamp is TS;Starting current maximum value I* is IM/IW, value about 9-10 it Between;The starting current maximum value time is TM, TM=0.
It is illustrated in figure 3 the start-up course current waveform of the resistive loads such as resistance furnace.The resistive loads such as resistance furnace are logical Frequently with lectrothermal alloy wires such as nickel chromium triangle, ferrum-chromium-aluminums, common feature is that resistance temperature correction factor is small, resistance value stabilization.With board For number for the nichrome wire of Cr20Ni80, resistance correction factor at 1000 DEG C is 1.014, i.e., 1000 DEG C whens are opposite When 20 DEG C, the trade mark is that the nichrome wire resistance of Cr20Ni80 only increases by 1.4%.Therefore, the resistive loads such as resistance furnace exist It is powered and enters stable state, the start-up course time T of the resistive loads such as resistance furnace when startingS=0;Starting current maximum value I*=1;Starting current maximum value time TM=0.
It is illustrated in figure 4 the start-up course current waveform of monophase machine class load.The load of monophase machine class both has inductance Property load characteristic, and have counter electromotive force load characteristic.Startup time, due to the effect of inductance, the starting current of Startup time It is 0;Subsequent electric current rises rapidly, before counter electromotive force of motor is not set up, reaches current peak IM;Hereafter, motor speed increases Add, motor load electric current gradually reduces, until entering stable state.In Fig. 4, the start-up course time of monophase machine class load is TS;Starting current maximum value I* is IM/IW;The starting current maximum value time is TM
It is illustrated in figure 5 computer and the start-up course current waveform of Switching Power Supply class load.Computer and Switching Power Supply Class load will produce a prodigious surge current because of the influence to capacitor charging, in startup moment, and peak value can reach surely State current effective value IWSeveral times to more than ten times, the time be 1 to 2 power frequency period.In Fig. 5, computer and Switching Power Supply class are negative The start-up course time of load is TS, about 1 to 2 power frequency period;Starting current maximum value I* is IM/IW;When starting current maximum value Between be TM=0.
Obtaining the method for starting current feature of electric appliance is:
Before appliance starting, when load current value is 0 (being not keyed up) or very little (being in standby), message processing module 102 start to carry out continuous sampling to load current;When the load current value virtual value that sampling obtains starts to be more than 0 or opens When beginning to be more than the standby current of electric appliance, that is, judge that electric appliance has been started up, it is T to record the moment0.With a smaller non-negative threshold Value ε distinguishes the load current value before and after appliance starting, when ε values are especially small, for example, when ε value 1mA, the identification dress It sets and does not consider ideal case, that is, it is also the starting state of electric appliance to think standby;When ε values are smaller but more than electric appliance standby current When, for example, when ε value 20mA, the standby mode of electric appliance can be considered inactive state by the identification device, but also can simultaneously The especially small electric appliance of Partial Power cause leakage to identify.
Message processing module 102 carries out continuous sampling to load current, and using power frequency period as unit computational load electric current Virtual value simultaneously preserves;Continuous plus is most after electric appliance has been started up, and continuous sampling reaches N number of power frequency period, while sampling The average value I of the load current virtual value of nearly N number of power frequency periodV;Message processing module 102 to nearest N number of power frequency period within The load current virtual value of each power frequency period is compared with the average value of the load current virtual value of N number of power frequency period, When error (or fluctuation) amplitude is respectively less than the relative error range E set, judgement electric appliance load enters stable state, the nearest N The initial time of a power frequency period is the finish time of start-up course, and it is T to record the moment1
The average value of load current virtual value within nearest N number of power frequency period is effective as electric appliance load steady-state current Value IW;Electric appliance is started into Startup time T0To nearest N number of power frequency period initial time T1Between time as the start-up course time TS;By T0To T1Within the moment where the maximum power frequency period of load current virtual value be recorded as T2, by T0To T2Between time As starting current maximum value time TM;By T2The load current virtual value of place power frequency period has with electric appliance load steady-state current Valid value IWBetween ratio as starting current maximum value I*.
Due to not knowing electric appliance load steady-state current virtual value I in advanceW, therefore, by N number of power frequency period, i.e., one section continues Time TPWithin load current virtual value of fluctuation range when being less than the relative error range E of setting average value it is negative as electric appliance Carry steady-state current virtual value IW.Since the start-up course of ordinary appliances load is very fast, so, TPValue range be 1-10s, allusion quotation Type value is 2s, and the value range of corresponding power frequency period quantity N is 50-500, and the typical value of N is 100.It is described to miss relatively The value range of poor range E is 2%-20%, and the typical value of E is 10%.
The input feature vector of assembled classifier further includes the load current spectrum signature of electric appliance.The load current frequency spectrum of electric appliance is special Sign controls information acquisition module 101 by message processing module 102, is obtained by following steps:
Step 1: after electric appliance load enters stable state, the steady state current signals of electric appliance load are obtained, and are converted For corresponding steady-state current digital signal.
Step 2: carrying out Fourier transform to steady-state current digital signal, load current spectral characteristic is obtained.To ensure Fu Vertical leaf transformation is smoothed out, and the steady state current signals of electric appliance load are obtained aforementioned, and is converted into corresponding stable state electricity During streaming digital signal, the accuracy and speed of A/D converter needs the requirement for meeting Fourier transform, and sample frequency can be with It is set as 10kHz or other numerical value;Message processing module 102 carries out FFT fortune to collected steady-state current digital signal It calculates, calculates its frequency spectrum.
Step 3: using the nth harmonic signal relative magnitude in load current spectral characteristic as load current spectrum signature, Wherein, n=1,2 ..., M;When forming the input feature value of assembled classifier, nth harmonic signal relative magnitude is special in input According to 1,2 in sign vector ..., the sequence of M is arranged in order.Since load current spectral characteristic is mainly made of odd harmonic, remove Outside a small number of electrical equipments, even-order harmonic component is almost 0, accordingly it is also possible to be n by overtone order in load current spectral characteristic Secondary odd harmonic signal relative magnitude is sequentially used as load current spectrum signature, wherein n=1,3 ..., M.1 time when n=1 Harmonic wave is fundamental frequency.The harmonic signal relative magnitude is harmonic signal amplitude and electric appliance load steady-state current virtual value IW's Ratio.The M indicates harmonic wave highest number, and under normal circumstances, M is more than or equal to 3.
The input feature vector of assembled classifier further includes the fundamental voltage current and phase difference of electric appliance.Fundamental voltage current and phase difference Resistive, capacitive character, inductive load can be distinguished, it can also be to general inductive load and big inductive load It distinguishes.The fundamental voltage current and phase difference of electric appliance by message processing module 102 control information acquisition module 101, by with Lower step obtains:
Step 1., after electric appliance load enters stable state, the synchronous steady state voltage signal for obtaining electric appliance load, stable state electricity Signal is flowed, and is converted into corresponding steady state voltage digital signal, steady-state current digital signal.
2., to steady state voltage digital signal, steady-state current digital signal step carries out digital filtering respectively, extract fundamental wave Voltage signal, fundamental current signal.
Step 3., analysis calculate the phase difference between fundamental voltage signal and fundamental current signal, by fundamental voltage signal Fundamental voltage current and phase difference of the phase difference as electric appliance between fundamental current signal.
Step 2. in digital filtering carried out respectively to steady state voltage digital signal, steady-state current digital signal, number filter Wave algorithm can select the digital filter algorithms such as Kalman filtering method, Wavelet Transform, Wiener Filter Method, adaptive-filtering.
In assembled classifier, decision tree classifier is main grader, and Bayes classifier is auxiliary grader.Assembled classification The input feature vector of device includes starting current feature above-mentioned and load current spectrum signature, and the input feature vector of assembled classifier is simultaneously As the input feature vector of decision tree classifier and the input feature vector of Bayes classifier.
It is illustrated in figure 6 electric appliance sorting technique flow chart, the electric appliance sorting technique comprises the concrete steps that:
Step A, appliance starting is waited for;
Step B, it acquires appliance starting current data and preserves, until appliance starting process terminates;
Step C, the appliance starting current data of analysis acquisition, obtains the starting current feature of electric appliance;
Voltage when step D, acquiring electric appliance steady operation and preserves current data;
Step E, the voltage when electric appliance steady operation of analysis acquisition, current data, obtain the load current frequency spectrum of electric appliance Feature, fundamental voltage current and phase difference;
Step F, using starting current feature, load current spectrum signature, fundamental voltage current and phase difference as assembled classification The input feature vector of device;Assembled classifier carries out appliance type identification;
Step G, appliance type recognition result is exported.
The assembled classifier, which carries out appliance type, which knows method for distinguishing, is:When Main classification device successfully realizes that appliance type is known Not, i.e., Main classification device output recognition result be unique appliance type, i.e., in recognition result unique appliance type for be when, The appliance type that Main classification device is identified is as the appliance type recognition result of assembled classifier;When Main classification device fails to realize electricity Device type identification, and the recognition result of Main classification device be 2 kinds either have in two or more appliance type, that is, recognition result 2 kinds or Two or more appliance type is when being, in 2 kinds that Main classification device is exported or two or more appliance type recognition result, auxiliary point Appliance type recognition result of the highest appliance type of probability as assembled classifier in the output of class device;When Main classification device fails reality Existing appliance type identification, and fail to provide the appliance type of identification in the recognition result of Main classification device, i.e., do not have in recognition result Appliance type is when being, the highest appliance type of probability is known as the appliance type of assembled classifier during subsidiary classification device is exported Other result.
By taking a simple embodiment 1 as an example, method for distinguishing is known to illustrate that assembled classifier carries out appliance type.Equipped with one A assembled classifier, input feature vector areWherein, TSIt is to open Dynamic process time, unit is ms;I* is starting current maximum value;TMIt it is the starting current maximum value time, unit is ms;A1、A2、 A3、A4、A5For the 1-5 rd harmonic signal relative magnitudes in load current spectral characteristic,For the fundamental voltage electric current of power load Phase difference.The output of assembled classifier is { B1, B2, B3, B4, B1、B2、B3、B4Assembled classifier is respectively represented to incandescent lamp, electricity Hinder the recognition result output of stove, hair-dryer, computer, recognition result B1、B2、B3、B4Value be two-value classification marker.It is main The input feature vector of grader is also Its output is { F1, F2, F3, F4, F1、F2、F3、F4It respectively represents Main classification device to export the recognition result of incandescent lamp, resistance furnace, hair-dryer, computer, identification As a result F1、F2、F3、F4Value be also two-value classification marker.The input feature vector of subsidiary classification device is similarlyIts output is { P (y1| x), P (y2| x), P (y3| x), P (y4| X) }, P (y1|x)、P(y2|x)、P(y3|x)、P(y4| x) it is the posterior probability for assisting grader output, P (y1|x)、P(y2|x)、P (y3|x)、P(y4| the mutual size between x) shows that the current input feature of subsidiary classification device indicates that identified electric appliance belongs to white The possibility size of vehement lamp, resistance furnace, hair-dryer, computer.
In embodiment 1, B1、B2、B3、B4Classification marker and F1、F2、F3、F4Classification marker take 1,0.Classification marker When being 1, corresponding appliance type is matched with current input feature, for the recognition result of confirmation, corresponding appliance type in other words Recognition result is yes;When classification marker is 0, corresponding appliance type is mismatched with input feature vector, is failed as the identification confirmed As a result, corresponding appliance type recognition result is no in other words.
In embodiment 1, if the recognition result classification marker of certain Main classification device is F1F2F3F4=0100, then it is assumed that Main classification device successfully realizes that appliance type identifies, therefore, does not consider the recognition result of subsidiary classification device, directly enables B1B2B3B4= 0100, i.e. the recognition result of assembled classifier is:Identified electric appliance is resistance furnace.
In embodiment 1, if the recognition result classification marker of certain Main classification device is F1F2F3F4=1010, then it is assumed that Main classification device fails to realize appliance type identification, and the recognition result of Main classification device is 2 kinds or two or more appliance type;Again If the recognition result of subsidiary classification device meets P (y at this time1|x)<P(y3| x), then enable B1B2B3B4=0010, i.e. assembled classifier Recognition result is:Identified electric appliance is hair-dryer.
In embodiment 1, if the recognition result classification marker of certain Main classification device is F1F2F3F4=0000, then it is assumed that Main classification device fails to realize appliance type identification, and fails to provide the appliance type of identification in the recognition result of Main classification device;Again If the recognition result of subsidiary classification device meets P (y at this time1|x)>P(y2| x) and P (y1|x)>P(y3| x) and P (y1|x)>P(y4| X), then B is enabled1B2B3B4=1000, i.e. the recognition result of assembled classifier is:Identified electric appliance is incandescent lamp.
The recognition result classification marker of assembled classifier, Main classification device can also use other schemes, for example, using respectively Classification marker 1, -1 or 0,1, either -1,1 and other schemes come indicate corresponding electric appliance recognition result be yes, it is no. Assembled classifier can be identical with the classification marker scheme of Main classification device, can not also be identical.
Can also include electric appliance load steady-state current virtual value I in the input feature vector of the assembled classifierW.For example, having 2 kinds of different electric appliances, electric iron and resistance furnace need to identify, electric iron, resistance furnace are all pure resistor loads, and all has resistance Temperature correction coefficient is small, the common feature of resistance value stabilization.Therefore, starting current feature above-mentioned and load current are relied on merely Spectrum signature, electric appliance fundamental voltage current and phase difference feature they can not be distinguished.It is negative to increase electric appliance in input feature vector Carry steady-state current virtual value IWAfterwards, electric iron power is small, electric appliance load steady-state current virtual value IWIt is small;Resistance furnace power is big, electric appliance Load steady state current effective value IWGreatly, feature is different, and assembled classifier can carry out and complete to identify.
Subsidiary classification device is Bayes classifier.It can select NBC graders (Naive Bayes Classifier), TAN classification Three kinds of Bayes classifiers such as device (crown pruning), BAN graders (Bayes classifier of enhancing) it In it is a kind of be used as subsidiary classification device.
Embodiment 2 selects NBC graders as subsidiary classification device.Naive Bayes Classification is defined as follows:
(1) x={ a are set1, a2..., amIt is an item to be sorted, and the characteristic attribute that each a is x;
(2) category set C={ y are had1, y2..., yn};
(3) P (y are calculated1| x), P (y2| x) ..., P (yn|x);
(4) if P (yk| x)=max { P (y1| x), P (y2| x) ..., P (yn| x) }, then x ∈ yk
Calculate the (3) walk in the specific method of each conditional probability be:
1. it is training sample set to find the item collection cooperation to be sorted classified known to one;
2. statistics obtain it is of all categories under each characteristic attribute conditional probability estimation;
P(a1|y1), P (a2|y1) ..., P (am|y1);
P(a1|y2), P (a2|y2) ..., P (am|y2);
…;
P(a1|yn), P (a2|yn) ..., P (am|yn)。
3. according to Bayes' theorem, have:
Because denominator for all categories be constant, as long as therefore we molecule is maximized;Again because in simplicity Each characteristic attribute is conditional sampling in Bayes, so having:
In embodiment 2, the input feature vector of assembled classifier is { TS, I*, TM, A1, A3,IW, wherein TSIt is start-up course Time, unit are ms;I* is starting current maximum value;TMIt it is the starting current maximum value time, unit is ms;A1、A3For load electricity Flow 1,3 odd harmonic signal relative magnitude in spectral characteristic;For the fundamental voltage current and phase difference of electric appliance, unit is degree, And fundamental voltage is when being ahead of fundamental current,IWFor electric appliance load steady-state current virtual value, unit is ampere.It is required that knowing Other electric appliance classification is incandescent lamp, resistance furnace, electric fan, computer, electric iron.Enable the characteristic attribute of Naive Bayes Classifier Combine x={ a1, a2, a3, a4, a5, a6, a7In element and assembled classifier input feature vector set in element sequentially { TS, I*, TM, A1, A3,IWCorrespond;The output category set C={ y of Naive Bayes Classifier1, y2, y3, y4, y5Then divide It is not corresponded with electric appliance classification incandescent lamp, resistance furnace, electric fan, computer, electric iron.
The process of training NBC graders includes:
1, segmentation division is carried out to characteristic attribute, carries out sliding-model control.In embodiment 2, the characteristic attribute taken is discrete Change method is:
a1:{a1< 50,50≤a1≤ 1000, a1> 1000 };
a2:{a2< 7,7≤a2≤ 11, a2> 11 };
a3:{a3< 20,20≤a3≤ 300, a3> 300 };
a4:{a4< 0.7,0.7≤a4≤ 0.9, a4> 0.9 };
a5:{a5< 0.02,0.02≤a5≤ 0.05, a5> 0.05 };
a6:{a6< -6, -6≤a6≤ 18, a6> 18 };
a7:{a7< 0.45, a7≥0.45}。
2, multigroup sample is acquired as training sample to every electric appliances type, while calculates and exists per electric appliances type sample The ratio occupied in all appliance type samples, that is, calculate separately P (y1)、P(y2)、P(y3)、P(y4)、P(y5).When every class electricity When device acquires identical sample size, for example, the sample more than 100 groups is acquired per electric appliances, wherein random per electric appliances Select 100 groups of samples as training sample, other are then used as test sample, and total training sample is 500 groups, and is had
P(y1)=P (y2)=P (y3)=P (y4)=P (y5)=0.2.
3, the frequency (ratio) of each characteristic attribute segmentation under each class condition of training sample is calculated, statistics obtains all kinds of The conditional probability estimation of each characteristic attribute under other, i.e., statistics calculates respectively
P(a1< 50 | y1)、P(50≤α1≤1000|y1)、P(a1> 1000 | y1);
P(a1< 50 | y2)、P(50≤a1≤1000|y2)、P(a1> 1000 | y2);
…;
P(a1< 50 | y5)、P(50≤a1≤1000|y5)、P(a1> 1000 | y5);
P(a2< 7 | y1)、P(7≤a2≤11|y1)、P(a2> 11 | y1);
P(a2< 7 | y2)、P(7≤a2≤11|y2)、P(a2> 11 | y2);
…;
P(a2< 7 | y5)、P(7≤a2≤11|y5)、P(a2> 11 | y5);
P(a3< 20 | y1)、P(20≤a3≤300|y1)、P(a3> 300 | y1);
P(a3< 20 | y2)、P(20≤a3≤300|y2)、P(a3> 300 | y2);
…;
P(a3< 20 | y5)、P(20≤a3≤300|y5)、P(a3> 300 | y5);
P(a4< 0.7 | y1)、P(0.7≤a4≤0.9|y1)、P(a4> 0.9 | y1);
P(a4< 0.7 | y2)、P(0.7≤a4≤0.9|y2)、P(a4> 0.9 | y2);
…;
P(a4< 0.7 | y5)、P(0.7≤a4≤0.9|y5)、P(a4> 0.9 | y5);
P(a5< 0.02 | y1)、P(0.02≤a5≤0.05|y1)、P(a5> 0.05 | y1);
P(a5< 0.02 | y2)、P(0.02≤a5≤0.05|y2)、P(a5> 0.05 | y2);
P(a5< 0.02 | y5)、P(0.02≤a5≤0.05|y5)、P(a5> 0.05 | y5);
P(a6< -6 | y1)、P(-6≤a6≤18|y1)、P(a6> 18 | y1);
P(a6< -6 | y2)、P(-6≤a6≤18|y2)、P(a6> 18 | y2);
…;
P(a6< -6 | y5)、P(-6≤a6≤18|y5)、P(a6> 18 | y5);
P(a7< 0.45 | y1)、P(a7≥0.45|y1);
P(a7< 0.45 | y2)、P(a7≥0.45|y2);
…;
P(a7< 0.45 | y5)、P(a7≥0.45|y5)。
By above-mentioned step 1, step 2, step 3, NBC classifier trainings are completed.Wherein, step 1 to characteristic attribute into Row segmentation is divided by manually determining, when carrying out disperse segmentaly to each input feature vector, the quantity of segmentation is 2 sections or 2 sections More than, for example, in embodiment 2, feature a1-a6All it is divided into 3 sections, feature a7It is divided into 2 sections.Each feature is specifically divided into how many sections, Result after the selection of fragmentation threshold can test test sample according to the Bayes classifier after training is adjusted.Step 2, step 3 is calculated by the either computer of message processing module 102 and is completed.
It is using the method that Bayes classifier is classified in the present invention:
1, using the input feature vector of assembled classifier as the input feature vector of Bayes classifier.In example 2, it will combine Input feature vector set { the T of graderS, I*, TM, A1, A3,IWInput feature vector x as Bayes classifier, and have x= {a1, a2, a3, a4, a5, a6, a7}。
2, according to training obtain it is of all categories under each characteristic attribute conditional probability estimate, respectively determine each input feature vector Where the segmentation of attribute and determine its probability P (a to every electric appliances classification1|y1)~P (am|yn), wherein electric appliance category set For C={ y1, y2..., yn}.In embodiment 2, electric appliance category set C={ y1, y2, y3, y4, y5The corresponding electric appliance classification represented is Incandescent lamp, resistance furnace, electric fan, computer, electric iron determine P (a1|y1)~P (a7|y5) method be using training NBC point The conditional probability estimation of each characteristic attribute obtained during class device.
3, according to formula
Calculate the other posterior probability of each electric type.Because denominator P (x) is constant for all electric appliance classifications, P (x) is enabled =1 substitutes actual P (x) value, and the mutual size not influenced between each electric appliance classification posterior probability compares, has at this time
In embodiment 2, have
Trained Bayes classifier is tested using test sample, adjustment pair is decided whether according to test result The discretization method (adjusting number of fragments and threshold value) of input feature vector, re -training Bayes classifier.
Main classification device is decision tree classifier, and the algorithm of decision tree classifier can select ID3, C4.5, CART etc..Implement Example 2 is selected using ID3 decision tree classifiers as Main classification device.The several of ID3 decision tree classifiers are defined as follows:
If D is the division carried out to training tuple with classification, then the entropy of D is expressed as:
Wherein piIndicate the probability of i-th of classification appearance in entire training tuple (i.e. sample), it can be with belonging to such The quantity divided by training tuple elements total quantity of other element are as estimation.The practical significance expression of entropy is the class label of tuple in D Required average information.
Assuming that training tuple D is divided by attribute A, then A is to the expectation information that D is divided:
And information gain is the difference of the two:
Gain (A)=info (D)-infoA(D) (3)
ID3 algorithms calculate the ratio of profit increase of each attribute when needing division every time, then select the maximum attribute of ratio of profit increase Into line splitting.
Characteristic attribute discretization method may be used in training ID3 decision tree classifiers, can also use continuous characteristic attribute Potential disintegrating method.Its specific method is:All attributes are detected, the maximum attribute of information gain is selected to generate decision tree knot Point establishes branch by the different values of the attribute, then establishes decision tree node to subset recursive call this method of each branch Branch, until all subsets only include same category of data.A decision tree is finally obtained, it can be used for new Sample is classified.In example 2, multigroup sample is acquired to every electric appliances type, randomly selects part as training sample This, remaining is as test sample.
Characteristic attribute discretization method training ID3 decision tree classifiers process include:
1) feature differentiation is realized to each characteristic attribute.In embodiment 2, the feature differentiation method taken is:
a1:{a1< 50,50≤a1≤ 1000, a1> 1000 };
a2:{a2< 4, a2≥4};
a3:{a3< 30, a3≥30};
a4:{a4< 0.85, a4≥0.85};
a5:{a5< 0.1, a5≥0.05};
a6:{a6< -12, -12≤a6≤ 12, a6> 12 };
a7:{a7< 0.45, a7≥0.45}。
2) information gain of each attribute is calculated.In example 2, it is counted respectively according to formula (2) and formula (3) for training sample Calculate the information gain of 7 characteristic attributes.
3) division (decision) attribute and decision tree node of the attribute with maximum information gain as the secondary division are selected, Division is obtained as a result, establishing branch;If sample is all in same class, which becomes leaf, such is used in combination to mark.
4) on the basis of existing division result, recurrence calculates the Split Attribute of child node using abovementioned steps, establishes and divides Branch, finally obtains entire decision tree.
By above-mentioned step, the training of ID3 decision tree classifiers is completed.Wherein, step 1) is segmented characteristic attribute Feature differentiation is by manually determining, when carrying out disperse segmentaly to each input feature vector, the quantity of segmentation be 2 sections or 2 sections with On, for example, in embodiment 2, feature a1、a6It is divided into 3 sections, feature a2-a5、a7All it is divided into 2 sections.Each feature is specifically divided into how many The selection of section, fragmentation threshold can be adjusted the result after test sample test according to the decision tree classifier after training. Step 2) is completed to step 4) by the either computer of message processing module 102.
Continuously the process of the potential disintegrating method training ID3 decision tree classifiers of characteristic attribute includes:
I, the information gain of each attribute is calculated.First element in training sample D is sorted according to characteristic attribute, then each two phase The intermediate point of neighbors can regard potential split point as, since first potential split point, divide two set of D and calculating It is expected that information, the point with minimum expectation information is known as the best splitting point of this attribute, and information it is expected as this attribute Information it is expected.In example 2, it for training sample, finds out best splitting point and calculates separately 7 spies according to formula (2) and formula (3) Levy the information gain of attribute.
II, division (decision) attribute and decision tree knot of the attribute with maximum information gain as the secondary division are selected Point obtains division as a result, establishing branch;If sample is all in same class, which becomes leaf, such is used in combination to mark.
III, on the basis of existing division result, recurrence calculates the Split Attribute of child node using abovementioned steps, establishes and divides Branch, finally obtains entire decision tree.
In the training process of aforementioned decision tree, when all samples of given node belong to same class, terminate recursive procedure, Decision tree has built up.All samples of given node belong to same class, it may be possible to single electric type is other confirm as a result, It may be the negative decision of all appliance types.
In the training process of aforementioned decision tree classifier, when do not have remaining attribute can be used for further divide sample When, it also needs to terminate recursive procedure, but some subsets are not also pure collection at this time, i.e., the element in set is not belonging to same class Not;At this point it is possible to using characteristic attribute is increased, for example, increasing by 5 times, 7 times in load current spectral characteristic in example 2 Etc. odd harmonics signal relative magnitude as new characteristic attribute, re -training is carried out to decision tree.After training or again The final part subset of decision tree classifier after training is not pure collection, when the element in set is not belonging to same category, Do not use subset " majority voting " mode using the most classification of occurrence number in subset as this node classification, but directly will be sub For all categories of concentration as this node classification, i.e., the described decision tree classifier can export the other confirmation knot of a variety of electric types Fruit.
Main classification device is also an option that be made of multiple two classes output decision tree classifier, each two classes output decision tree point Class device, which corresponds to, identifies a kind of appliance type, knows respectively for example, 4 two classes output decision tree classifiers may be used in embodiment 1 5 two classes output decision tree classifiers may be used in embodiment 2 and know respectively for other incandescent lamp, resistance furnace, hair-dryer, computer Other incandescent lamp, resistance furnace, electric fan, computer, electric iron.Main classification device selects multiple two classes output decision tree classifiers common When composition, the input feature vector of all two classes output decision tree classifiers is the input feature vector of Main classification device, all training samples This exports the training sample of decision tree classifier as each two class.Main classification device selects multiple two classes to export decision tree classification When device collectively constitutes, each two classes output decision tree classifier only needs to be performed a kind of identification of appliance type, the instruction of decision tree White silk is relatively easy.After the training that some described two class export decision tree classifier, or increase characteristic attribute again After training, some subsets are not also pure collection, that is, have subset can't confirmation input attribute whether belong to two class output It is yes by the node definition where the subset when appliance type that decision tree classifier is identified, that is, two class is allowed to export decision Tree Classifier judges that the characteristic attribute this time inputted belongs to identified appliance type in this case.Due to Main classification at this time Device is made of multiple two classes output decision tree classifier, between each two classes output decision tree classifier independently of each other, therefore, to certain When one characteristic attribute is identified, the recognition result that Main classification device is possible to output is unique appliance type, or identification knot Fruit is 2 kinds and either two or more appliance type or fails to provide the appliance type of identification.

Claims (9)

1. a kind of electric appliance sorting technique, which is characterized in that using the combination point for including decision tree classifier and Bayes classifier Class device carries out appliance type identification;The input feature vector of the assembled classifier include the starting current feature of electric appliance, electric appliance it is negative Carry the fundamental voltage current and phase difference of current spectrum feature and electric appliance.
2. electric appliance sorting technique as described in claim 1, which is characterized in that in the assembled classifier, decision tree classifier For main grader, Bayes classifier is auxiliary grader.
3. electric appliance sorting technique as claimed in claim 2, which is characterized in that the assembled classifier carries out appliance type identification Method be:When Main classification device successfully realizes appliance type identification, the appliance type recognition result of Main classification device is combination point The recognition result of class device;It is identified when Main classification device fails realization appliance type, and the recognition result of Main classification device is 2 kinds or 2 Kind or more appliance type, in 2 kinds that Main classification device is exported or two or more appliance type recognition result, subsidiary classification device is defeated Go out appliance type recognition result of the highest appliance type of middle probability as assembled classifier;When Main classification device fails to realize electric appliance Type identification, and when failing to provide the appliance type of identification in the recognition result of Main classification device, will be general in the output of subsidiary classification device Appliance type recognition result of the highest appliance type of rate as assembled classifier.
4. electric appliance sorting technique as claimed in any one of claims 1-3, which is characterized in that the starting current feature includes Start-up course time, starting current maximum value, starting current maximum value time.
5. electric appliance sorting technique as claimed in any one of claims 1-3, which is characterized in that the load current spectrum signature It is prepared by the following:
Step 1: obtaining the steady state current signals of electric appliance load, and it is converted into corresponding steady-state current digital signal;
Step 2: carrying out Fourier transform to steady-state current digital signal, load current spectral characteristic is obtained;
Step 3: using the odd harmonic signal relative magnitude that overtone order in load current spectral characteristic is n times as load electricity Flow spectrum signature, wherein n=1,3 ..., M;The M indicates harmonic wave highest number and M is more than or equal to 3.
6. electric appliance sorting technique as claimed in claim 5, which is characterized in that the harmonic signal relative magnitude is harmonic signal The ratio of amplitude and electric appliance load steady-state current virtual value.
7. electric appliance sorting technique as claimed in claim 4, which is characterized in that the fundamental voltage current and phase difference of the electric appliance is logical Cross following methods acquisition:
Step 1., after electric appliance load enters stable state, the synchronous steady state voltage signal for obtaining electric appliance load, steady-state current letter Number, and it is converted into corresponding steady state voltage digital signal, steady-state current digital signal;
2., to steady state voltage digital signal, steady-state current digital signal step carries out digital filtering respectively, extract fundamental voltage Signal, fundamental current signal;
Step 3., analysis calculate the phase difference between fundamental voltage signal and fundamental current signal, by fundamental voltage signal and base Fundamental voltage current and phase difference of the phase difference as electric appliance between signal wave current.
8. electric appliance sorting technique as claimed in claim 7, which is characterized in that described after electric appliance load enters stable state Judge, is carried out according to the load current virtual value of each power frequency period within nearest N number of power frequency period;The value of the N Ranging from 50-500.
9. electric appliance sorting technique as claimed in claim 8, which is characterized in that continuous sampling is carried out to the load current of electric appliance, It as unit computational load current effective value and is preserved using power frequency period;Calculate the load current virtual value of N number of power frequency period recently Average value;The load current virtual value of each power frequency period within nearest N number of power frequency period and N number of power frequency period The average value of load current virtual value compares, and when fluctuating range is respectively less than the relative error range E set, judges electric appliance load Into stable state;The value range of the E is 2%-20%.
CN201810455735.1A 2016-04-08 2016-04-08 A kind of electric appliance sorting technique Pending CN108318770A (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201810455735.1A CN108318770A (en) 2016-04-08 2016-04-08 A kind of electric appliance sorting technique
CN201610213373.6A CN105866581B (en) 2016-04-08 2016-04-08 A kind of appliance type recognition methods

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810455735.1A CN108318770A (en) 2016-04-08 2016-04-08 A kind of electric appliance sorting technique

Related Parent Applications (1)

Application Number Title Priority Date Filing Date
CN201610213373.6A Division CN105866581B (en) 2016-04-08 2016-04-08 A kind of appliance type recognition methods

Publications (1)

Publication Number Publication Date
CN108318770A true CN108318770A (en) 2018-07-24

Family

ID=56636926

Family Applications (2)

Application Number Title Priority Date Filing Date
CN201810455735.1A Pending CN108318770A (en) 2016-04-08 2016-04-08 A kind of electric appliance sorting technique
CN201610213373.6A Active CN105866581B (en) 2016-04-08 2016-04-08 A kind of appliance type recognition methods

Family Applications After (1)

Application Number Title Priority Date Filing Date
CN201610213373.6A Active CN105866581B (en) 2016-04-08 2016-04-08 A kind of appliance type recognition methods

Country Status (1)

Country Link
CN (2) CN108318770A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109596917A (en) * 2018-12-07 2019-04-09 江苏智臻能源科技有限公司 A kind of non-load identification efficient detection method of registering one's residence

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107273920A (en) * 2017-05-27 2017-10-20 西安交通大学 A kind of non-intrusion type household electrical appliance recognition methods based on random forest
CN108224681A (en) * 2017-12-16 2018-06-29 广西电网有限责任公司电力科学研究院 Non-intrusion type starting of air conditioner detection method based on decision tree classifier
CN109030975B (en) * 2018-05-23 2020-02-18 北京航空航天大学 Electric appliance type inference method and device based on intelligent socket
CN109164327A (en) * 2018-10-11 2019-01-08 深圳华建电力工程设计有限公司 Electricity system electric appliance behavior discrimination method and device based on combined type criterion
CN109613362B (en) * 2018-12-14 2021-02-26 四川长虹电器股份有限公司 Non-invasive electric appliance quantity identification method

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2362769A (en) * 2000-05-26 2001-11-28 Nokia Mobile Phones Ltd Battery charging circuit in which power is supplied via a communications port
CN2731801Y (en) * 2004-02-16 2005-10-05 哈尔滨世纪旌旗科技有限责任公司 Diggings power supply management arrangement
KR20120064581A (en) * 2010-12-09 2012-06-19 한국전자통신연구원 Mehtod of classfying image and apparatus for the same
CN103378506B (en) * 2012-04-20 2015-10-28 Tcl集团股份有限公司 A kind of electrical source socket of identifiable design appliance type
CN102684270B (en) * 2012-05-31 2014-12-10 华为技术有限公司 Method for identifying type of universal serial bus (USB) chargers and USB device
CN103489005B (en) * 2013-09-30 2017-04-05 河海大学 A kind of Classification of High Resolution Satellite Images method based on multiple Classifiers Combination

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109596917A (en) * 2018-12-07 2019-04-09 江苏智臻能源科技有限公司 A kind of non-load identification efficient detection method of registering one's residence
WO2020114140A1 (en) * 2018-12-07 2020-06-11 江苏智臻能源科技有限公司 Efficient detection method for non-household load identification

Also Published As

Publication number Publication date
CN105866581B (en) 2018-06-26
CN105866581A (en) 2016-08-17

Similar Documents

Publication Publication Date Title
CN105866581B (en) A kind of appliance type recognition methods
CN105785187B (en) A kind of students&#39; dormitory electrical appliance type judgement method
CN105759148B (en) A kind of electrical appliance type judgement method
Dong et al. Non-intrusive signature extraction for major residential loads
CN105426860B (en) The method and apparatus of recognition of face
CN106909101B (en) A kind of non-intrusion type household electrical appliance sorter and method
Du et al. Electric load classification by binary voltage–current trajectory mapping
CN105913005A (en) Electric appliance load type intelligent identification method and device
CN106872824A (en) A kind of network load appliance type identification and the method and its device of different electrical equipment electricity statistics
CN106443244B (en) Electrical equipment type discrimination method and system
Ondel et al. A method to detect broken bars in induction machine using pattern recognition techniques
WO2013081717A2 (en) System and method employing a hierarchical load feature database to identify electric load types of different electric loads
CN109492667A (en) A kind of feature selecting discrimination method for non-intrusive electrical load monitoring
CN105913010A (en) Electric appliance type determination device
CN105891633A (en) Dormitory electric apparatus type determination device
CN105759149B (en) Students&#39; dormitory electrical appliance type detector
CN105913006A (en) Electricity load type identification method
CN105868790A (en) Electrical load type recognizer
CN105913009A (en) Electric appliance type identifier
CN104657422B (en) A kind of content issue intelligent method for classifying based on categorised decision tree
CN110135392A (en) A kind of electrical load kind identification method
CN105866580A (en) Electric appliance type determining apparatus
CN105891634A (en) Electric appliance type identification device
CN110991786A (en) 10kV static load model parameter identification method based on similar daily load curve
CN110197220A (en) A kind of electrical load starting operation recognition methods

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination