CN110514889A - A kind of method and system of non-intrusion type household electricity remained capacity - Google Patents
A kind of method and system of non-intrusion type household electricity remained capacity Download PDFInfo
- Publication number
- CN110514889A CN110514889A CN201910653275.8A CN201910653275A CN110514889A CN 110514889 A CN110514889 A CN 110514889A CN 201910653275 A CN201910653275 A CN 201910653275A CN 110514889 A CN110514889 A CN 110514889A
- Authority
- CN
- China
- Prior art keywords
- load
- power
- unit
- real
- initial data
- 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
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R19/00—Arrangements for measuring currents or voltages or for indicating presence or sign thereof
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R21/00—Arrangements for measuring electric power or power factor
- G01R21/001—Measuring real or reactive component; Measuring apparent energy
- G01R21/002—Measuring real component
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R22/00—Arrangements for measuring time integral of electric power or current, e.g. electricity meters
- G01R22/06—Arrangements for measuring time integral of electric power or current, e.g. electricity meters by electronic methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
- G06F2218/04—Denoising
- G06F2218/06—Denoising by applying a scale-space analysis, e.g. using wavelet analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Abstract
The invention discloses a kind of method and system of non-intrusion type household electricity remained capacity, wherein the method for non-intrusion type household electricity remained capacity includes acquiring the real-time voltage electric current initial data of the various power loads in target user family inlet；Active power sequence, and given threshold are calculated according to the real-time voltage electric current initial data, calculates the power swing value in active power sequence sliding window, and judge whether the power swing value is greater than the given threshold；The load signal transient state and stable state various dimensions feature for extracting the real-time voltage electric current initial data are input to based on progress load identification and decomposition, output category result in the trained classifier of load characteristic java standard library.It realizes and the electric power and working condition of each or every electric appliances in subscriber household is specified using DAG_SVM algorithm classification by acquiring and analyzing entrance total current, information of voltage, obtain the electricity consumption rule of resident.
Description
Technical field
The present invention relates to intelligent power and load identification technology fields more particularly to a kind of non-intrusion type household electricity to load
The method and system of identification.
Background technique
Intelligent power is one of the important link of strong smart grid, is the core of interactive service system, key technology
It is mainly reflected in advanced measurement system (advanced metering infrastructure, AMI) standard, system and terminal skill
Art, intelligent power two-way interaction operational mode and support technology and user power utilization environment and 3 sides that influence each other with power mode
Face.Load monitoring is the first step for realizing smart grid as one of most important component part of AMI, it is necessary to break through electricity at present
Table can only read electricity consumption total amount automatically, and cannot analyse in depth user's internal load ingredient, the limited equal bottles of the information on load of acquisition
Neck, to improve power information acquisition system and intelligent electric power utilization system, support two-way interaction service and intelligence use can be serviced.Therefore such as
What judges the electric power of each or every electric appliances in subscriber household by acquiring and analyzing the information such as entrance total current, voltage
And working condition, to show that the problem of the electricity consumption rule of resident is urgently to be resolved.
Summary of the invention
The purpose of the present invention is to provide a kind of method and system of non-intrusion type household electricity remained capacity, improve training
Efficiency reduces development cost, and can combine load identification main website identification and update unknown load, judge in subscriber household each or
The electric power and working condition of every electric appliances, to obtain the electricity consumption rule of resident.
In a first aspect, the embodiment of the invention provides a kind of methods of non-intrusion type household electricity remained capacity, comprising: adopt
Collect the voltage and current data of the various power loads of residential households, the multidimensional characteristic for extracting various power loads is trained, and is established
Load characteristic java standard library, wherein the multidimensional characteristic includes that active power, reactive power, current total harmonic distortion rate, electric current are humorous
When wave containing ratio, transient current waveform, transient current peak value, rush of current multiple, power rush multiple or transient process continue
Between one of or it is a variety of；
Acquire the real-time voltage electric current initial data of the various power loads in target user family inlet；
Active power sequence, and given threshold are calculated according to the real-time voltage electric current initial data, calculates active power
Power swing value in sequence sliding window, and judge whether the power swing value is greater than the given threshold；
The load signal transient state and stable state various dimensions feature for extracting the real-time voltage electric current initial data, are input to and are based on
Load identification and decomposition, output category result are carried out in the trained classifier of load characteristic java standard library；Wherein, described point
Class device is classified based on DAG_SVM algorithm.
In one embodiment, before calculating active power sequence according to the real-time voltage electric current initial data, institute
State method further include:
It carries out smooth to the real-time voltage electric current initial data using Wavelet Transform Threshold method and is handled except making an uproar.
In one embodiment, active power sequence is calculated according to the real-time voltage electric current initial data, and sets threshold
Value calculates the power swing value in active power sequence sliding window, and judges whether the power swing value is greater than the setting threshold
Value, wherein judge whether the power swing value is greater than the given threshold, comprising:
If the power swing value is greater than the given threshold, the load of the real-time voltage electric current initial data is extracted
Signal transient state and stable state various dimensions feature are input to based on carrying out load in the trained classifier of load characteristic java standard library
Identification and decomposition, output category result；
If the power swing value is less than or equal to the given threshold, the real-time of subscriber household inlet is resurveyed
Voltage and current initial data.
In one embodiment, before output category result, the method also includes:
If the classifier can recognize that target load, the opening time and end time of the target load are recorded,
And add up electric flux when target load operation, output display result；
If the classifier not can recognize that target load, the characteristic value for extracting target load stores and is uploaded to main website
Load characteristic java standard library is updated after being identified.
Second aspect, the embodiment of the invention provides a kind of identifying systems, including executing any one of above-mentioned first aspect
The unit of the method for the non-intrusion type household electricity remained capacity, specifically, the identifying system include data sampling unit,
Incident detection unit, load characteristic library unit, load recognition unit, storage unit, load recognition result display unit and load
Identify main website, the incident detection unit, load characteristic library unit, the storage unit, load identification main website are and institute
The electric connection of load recognition unit is stated, the data sampling unit and time probe unit are electrically connected, the load identification knot
Fruit display unit is electrically connected；Wherein,
The data sampling unit, for acquiring the real-time voltage and electric current initial data of subscriber household inlet, and will
The initial data of acquired voltage and current is sent to the incident detection unit；
The incident detection unit, for carrying out received primary voltage current data at wavelet threshold-value filter
Reason, with filtered voltage and current data calculate power sequence, by sliding window detection method detect power sequence mutation value whether
More than set threshold value；
The load characteristic library unit is stored in load for carrying out feature extraction and training to resident's common load equipment
Characteristic standard library；
The load recognition unit, the primary voltage current data for will receive are handled and are calculated, calculated
The multidimensional characteristic information of remained capacity needed for method, and be directed into classifier and decomposed and identified；
The storage unit, the multidimensional for storage section primary voltage current sampling data, after processing and calculating are special
Levy data and load recognition result data；
The load recognition result display unit, handles for the result to load recognition unit, obtains each
Starting time, shut-in time and the power consumption within the corresponding period of electrical equipment, and classification is carried out to every electric appliances and is shown；
The load identifies main website, the update for identification and load characteristic java standard library to unknown load.
The method and system of a kind of non-intrusion type household electricity remained capacity of the invention, by extracting various power loads
Multidimensional characteristic be trained, establish load characteristic java standard library；Acquire the reality of the various power loads in target user family inlet
When voltage and current initial data, and calculate active power sequence and calculate active power sequence sliding window in power swing value, sentence
Whether the power swing value of breaking is greater than the given threshold, and the load signal for extracting the real-time voltage electric current initial data is temporary
State and stable state various dimensions feature, be input to based in the trained classifier of load characteristic java standard library carry out load identification and
It decomposes, output category result；It realizes by acquiring and analysis entrance total current, information of voltage, using DAG_SVM algorithm classification,
The electric power and working condition for specifying each or every electric appliances in subscriber household obtain the electricity consumption rule of resident；It is described simultaneously
The method of non-intrusion type household electricity remained capacity have that algorithm complexity is high, fast convergence rate, hardware resource requirements are small and
The higher advantage of recognition accuracy；The identifying system combines local identification and main website identification and one, can identify unknown negative
Lotus and the local load characteristic java standard library of update, improve the recognition accuracy to the unknown load of residential households.
Detailed description of the invention
It, below will be to specific in order to illustrate more clearly of the specific embodiment of the invention or technical solution in the prior art
Embodiment or attached drawing needed to be used in the description of the prior art are briefly described.In all the appended drawings, similar element
Or part is generally identified by similar appended drawing reference.In attached drawing, each element or part might not be drawn according to actual ratio.
Fig. 1 is the overall flow schematic diagram of the method for non-intrusion type household electricity remained capacity of the present invention；
Fig. 2 is the idiographic flow schematic diagram of the method for non-intrusion type household electricity remained capacity；
Fig. 3 is the DAG_SVM algorithm classification schematic diagram of 5 class household electrical appliance；
Fig. 4 is the structural schematic diagram of identifying system；
In figure: 100- identifying system, 10- data sampling unit, 20- incident detection unit, 30- load characteristic library unit,
40- load recognition unit, 50- storage unit, 60- load recognition result display unit, 70- load identify main website.
Specific embodiment
The embodiment of the present invention is described below in detail, examples of the embodiments are shown in the accompanying drawings, wherein from beginning to end
Same or similar label indicates same or similar element or element with the same or similar functions.Below with reference to attached
The embodiment of figure description is exemplary, it is intended to is used to explain the present invention, and is not considered as limiting the invention.
Fig. 1 and Fig. 2 are please referred to, Fig. 1 and Fig. 2 are that a kind of non-intrusion type household electricity load provided in an embodiment of the present invention is known
The flow diagram of method for distinguishing.Specifically, as depicted in figs. 1 and 2, the method for the non-intrusion type household electricity remained capacity
It may comprise steps of:
S101, the voltage and current data for acquiring the various power loads of residential households, the multidimensional for extracting various power loads are special
Sign is trained, and establishes load characteristic java standard library.
In the embodiment of the present invention, the voltage and current data of the various power loads first in the life of acquisition everyday home, such as
Then it is negative to extract each for refrigerator, air-conditioning, television set, computer, washing machine, electromagnetic oven, humidifier, water heater, soy bean milk making machine etc.
The multidimensional characteristic of load is trained, and establishes load characteristic java standard library.The multidimensional characteristic includes active power, reactive power, electricity
Flow total harmonic distortion factor, current harmonics containing ratio, transient current waveform, transient current peak value, rush of current multiple, power rush
One of multiple or transient process duration are a variety of, and the active power is to be actually sent out or consume in the unit time
Alternating current energy is the mean power in the period；The reactive power is electric field or magnetic field in the alternating current circuit with reactance
Energy is absorbed from power supply in the portion of time in a period, another part time then releases energy, average in whole cycle
Power is zero, but energy ceaselessly exchanges between power supply and reactance component (capacitor, inductance).The maximum value of exchange rate is
" reactive power ".In single phase alternating current circuit, value is equal to phase angle between voltage effective value, current effective value and voltage and electric current
The product of sinusoidal three.The current total harmonic distortion rate is the ratio between total harmonic current virtual value Ih and fundamental current virtual value.Often
It is expressed as a percentage.Current harmonics total harmonic distortion THDi=√ (I2*I2+I3*I3+...+In*In) * 100%/I1, In in formula
For n-th harmonic current effective value, I1 is fundamental current virtual value.Total harmonic distortion factor refers to the harmonic wave in periodical of ac
The percentage of the ratio between the root-mean-square valve of the root-mean-square valve of content and fundametal compoment.The current harmonics containing ratio is periodically exchange
The root-mean-square valve of harmonic content in amount and the ratio between the root-mean-square valve of its fundametal compoment, current total harmonic distortion rate THDI=IH/
I1 × 100%, in formula: IH is harmonic current content.The transient process duration is circuit from a stable state to another
The time of one stable state process experienced.Detailed process is S201: various electric appliance load voltage current samples；Load is more
Dimensional feature extracts and model training；Establish load characteristic java standard library.
S102, the real-time voltage electric current initial data for acquiring the various power loads in target user family inlet, according to institute
It states real-time voltage electric current initial data and calculates active power sequence, and given threshold, calculate the function in active power sequence sliding window
Rate undulating value, and judge whether the power swing value is greater than the given threshold.
In the embodiment of the present invention, the real-time voltage electric current original number of the various power loads in target user family inlet is acquired
According to, and carry out smooth to the real-time voltage electric current initial data using Wavelet Transform Threshold method and handled except making an uproar.Usual situation
Under, the signal of acquisition has certain noise, and in most cases this noise is white Gaussian noise, signal polluted by noise etc.
Noise is added in clean signal.Wavelet transformation is a kind of analysis method of the Time And Frequency of signal, and there is differentiate to divide more for it
The characteristics of analysis, and all there is the ability of characterization signal local feature in two domain of time-frequency, be a kind of window size immobilize but
Its shape is changeable, the Time-Frequency Localization analysis method that time window and frequency window can change.I.e. low frequency part have compared with
Low temporal resolution and higher frequency resolution, in high frequency section temporal resolution with higher and lower frequency point
Resolution is well suited for the signal of analysis non-stationary and extracts the local feature of signal, so wavelet transformation is known as analysis processing
The microscope of signal.Wavelet Transform Threshold method denoising process is specially since signal is spatially (or time-domain) to have centainly
It is successional, therefore in wavelet field, its modulus value of wavelet coefficient caused by useful signal is often larger；And white Gaussian noise is in sky
Between upper (or time-domain) be no successional, therefore noise passes through wavelet transformation, still behaves as in small echo threshold very strong
Randomness is usually still considered Gauss white noise.In wavelet field, the corresponding coefficient of useful signal is very big, and the corresponding system of noise
Number very little.Noise still meets the distribution of Gauss white noise in the corresponding coefficient of wavelet field.If in wavelet field, the wavelet coefficient pair of noise
The variance answered is sigma, then most (99.99%) noise coefficients are all located at [- 3* according to the characteristic of Gaussian Profile
Sigma, 3*sigma] in section (Chebyshev inequality, 3sigma criterion).Therefore, as long as by section [- 3*sigma, 3*
Sigma] in coefficient zero setting (here it is the effects of common hard threshold function), can utmostly inhibit noise.It will be through
Wavelet coefficient reconstruct after crossing threshold process, so that it may the signal after being denoised.Then to the voltage and current after filtering and noise reduction
Data calculate active power sequence, and whether the mutation value that power sequence is detected by sliding window detection method is more than set threshold value,
To determine whether the event that load equipment investment, excision or state change occurs；Wherein sliding window detection method is each by limiting
Institute's received maximum number of cell of energy controls portfolio in a time window.Whether specially judge the power swing value
It is original to extract the real-time voltage electric current if the power swing value is greater than the given threshold greater than the given threshold
The load signal transient state and stable state various dimensions feature of data, are input to based on the trained classifier of load characteristic java standard library
Middle progress load identification and decomposition, output category result；If the power swing value is less than or equal to the given threshold, weigh
The real-time voltage electric current initial data of new acquisition subscriber household inlet.Detailed process is S202: real-time residential households inlet
Voltage and current sampling；Data inhibit noise and rejecting abnormalities point except making an uproar and smoothing processing；Whether judge the power swing value
Electrical equipment investment or excision are judged whether there is greater than the given threshold.
S103, the load signal transient state and stable state various dimensions feature for extracting the real-time voltage electric current initial data, input
To based on progress load identification and decomposition, output category result in the trained classifier of load characteristic java standard library.
In the embodiment of the present invention, the classifier is classified based on DAG_SVM algorithm, and DAG_SVM algorithm is to realize
The method for how separating the data of user's needs further calculated out further according to predefined label or output for data conversion.
Now for example, Fig. 3 is the DAG_SVM algorithm classification schematic diagram of 5 class household electrical appliance, A, B, C, D, E have respectively represented air-conditioning, electricity
Hot-water bottle, electric fan, micro-wave oven and lighting apparatus, specific algorithm step are as follows:
S301, five kinds of electric appliances are input in (A, E) classifier sample is divided into ABCD and BCDE；
S302, again by the classification results of S201 by (A, D) and (B, E) classifier, obtain classification results ABC, BCD and
CDE；
S303, the classification results of S202 are directed respectively into (A, C), (B, D) and (C, E) again, obtain classification results AB, BC,
CD and DE；
S304, the classification results of S203 are finally passed through into (A, B), (B, C), (C, D) and (D, E) respectively, are obtained final
Classification results A, B, C, D and E.
Concrete principle is as follows: finding a hyperplane y=wTX+b classifies sample, wherein class intervalIt is maximum
Be optimal hyperlane, specific algorithm may be expressed as:
Household electrical appliance type Z=[z to be sorted1..., zn], every electric appliances type has k dimensional feature X=[x1...,
xk], optimization classification, objective function are carried out using the feature of every electric appliances are as follows:
s.t.yi(ωTx+b)≥1-ξi, i=1 ..., m, ξi≥0
Wherein C is penalty factor, ξiFor slack variable, yiFor classification results.
It is converted using its dual problem, and is solved by method of Lagrange multipliers, detailed process are as follows:
Wherein ui>=0, ai>=0 is Lagrange multiplier.Therefore the optimal solution of (ω, b, ξ), can be by enabling respective local derviation
It is obtained equal to zero:
Since most household electrical appliance type is all Nonlinear separability, it is therefore desirable to choose kernel function appropriate, we
The optional kernel function of method has RBF core or sigmoid core, sample is mapped to more higher-dimension from luv space, so that it becomes line
Property can divide.It can finally identify and various household household electrical appliance of classifying.
If the classifier can recognize that target load, the opening time and end time of the target load are recorded,
And add up electric flux when target load operation, output display result；
If the classifier not can recognize that target load, the characteristic value for extracting target load stores and is uploaded to main website
Load characteristic java standard library is updated after being identified.Detailed process is S203: utilizing classifier, carries out load decomposition and identification；Point
Whether class device has result output, if so, result output display；If it is not, unknown load is then uploaded to main website.
Referring to Fig. 4, Fig. 4 is a kind of structural schematic diagram of identifying system 100 provided in an embodiment of the present invention.Such as Fig. 4 institute
Show, the identifying system 100 includes the non-intrusion type household electricity remained capacity for executing any one of above-mentioned first aspect
The unit of method, specifically, the identifying system 100 includes data sampling unit 10, incident detection unit 20, load characteristic library
Unit 30, load recognition unit 40, storage unit 50, load recognition result display unit 60 and load identify main website 70, described
Incident detection unit 20, load characteristic library unit 30, the storage unit 50, the load identification main website 70 with the load
Recognition unit 40 is electrically connected, and the data sampling unit 10 is electrically connected with time probe unit, the load recognition result
Display unit 60 is electrically connected；Wherein,
The data sampling unit 10, for acquiring the real-time voltage and electric current initial data of subscriber household inlet, and
The initial data of acquired voltage and current is sent to the incident detection unit 20；
The incident detection unit 20, for carrying out received primary voltage current data at wavelet threshold-value filter
Reason, with filtered voltage and current data calculate power sequence, by sliding window detection method detect power sequence mutation value whether
More than set threshold value；
The load characteristic library unit 30, for carrying out feature extraction and training to resident's common load equipment, deposit is negative
He Tezhengbiaozhunku；
The load recognition unit 40, the primary voltage current data for will receive are handled and are calculated, obtained
The multidimensional characteristic information of remained capacity needed for algorithm, and be directed into classifier and decomposed and identified；
The storage unit 50, for storage section primary voltage current sampling data, the multidimensional after processing and calculating
Characteristic and load recognition result data；
The load recognition result display unit 60, is handled for the result to load recognition unit 40, is obtained every
Starting time, shut-in time and the power consumption within the corresponding period of one electrical equipment, and it is aobvious to carry out classification to every electric appliances
Show；
The load identifies main website 70, the update for identification and load characteristic java standard library to unknown load.
Detailed process are as follows: the multidimensional characteristic that the load characteristic library unit 30 extracts life family's common load is instructed
Practice, establishes load characteristic java standard library；The data sampling unit 10 acquires the real-time voltage and electricity of target user family inlet
Flow initial data；The incident detection unit 20 receives the real-time voltage and electric current initial data, with Wavelet Transform Threshold
Method noise-removed filtering, and power sequence is calculated to the voltage and current data after noise-removed filtering, power sequence is detected by sliding window detection method
Whether the mutation value of column is more than set threshold value, that is, judges whether to occur what load equipment investment, excision or state changed
Event；The load recognition unit 40 meets characteristic standard library with DAG_ in conjunction with what the load characteristic library unit 30 was established
SVM algorithm is classified, and identifies target load, then the classification of load recognition result display unit 60 shows that each electricity consumption is set
Standby starting time, shut-in time and the power consumption within the corresponding period, unidentified target load out, then the load identification is led
70 pairs of unknown loads of standing are identified and are updated load characteristic java standard library, the voltage and current in 50 storing process of storage unit
Sampled data, the multi-dimensional feature data after processing and calculating and load recognition result data.
In the various embodiments of the application, magnitude of the sequence numbers of the above procedures are not meant to the elder generation of execution sequence
Afterwards, the execution sequence of each process should be determined by its function and internal logic, the implementation process structure without coping with the embodiment of the present invention
At any restriction.
Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention., rather than its limitations；To the greatest extent
Pipe present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that: its according to
So be possible to modify the technical solutions described in the foregoing embodiments, or to some or all of the technical features into
Row equivalent replacement；And these are modified or replaceed, various embodiments of the present invention technology that it does not separate the essence of the corresponding technical solution
The range of scheme should all cover within the scope of the claims and the description of the invention.
Claims (5)
1. a kind of method of non-intrusion type household electricity remained capacity characterized by comprising
The voltage and current data of the various power loads of residential households are acquired, the multidimensional characteristic for extracting various power loads is instructed
Practice, establishes load characteristic java standard library, wherein the multidimensional characteristic includes active power, reactive power, current total harmonic distortion
Rate, current harmonics containing ratio, transient current waveform, transient current peak value, rush of current multiple, power rush multiple or transient state mistake
One of journey duration is a variety of；
Acquire the real-time voltage electric current initial data of the various power loads in target user family inlet；
Active power sequence, and given threshold are calculated according to the real-time voltage electric current initial data, calculates active power sequence
Power swing value in sliding window, and judge whether the power swing value is greater than the given threshold；
The load signal transient state and stable state various dimensions feature for extracting the real-time voltage electric current initial data are input to based on described
Load identification and decomposition, output category result are carried out in the trained classifier of load characteristic java standard library；Wherein, the classifier
It is to be classified based on DAG_SVM algorithm.
2. the method for non-intrusion type household electricity remained capacity as described in claim 1, which is characterized in that according to the reality
When voltage and current initial data calculate active power sequence before, the method also includes:
It carries out smooth to the real-time voltage electric current initial data using Wavelet Transform Threshold method and is handled except making an uproar.
3. the method for non-intrusion type household electricity remained capacity as described in claim 1, which is characterized in that according to described real-time
Voltage and current initial data calculates active power sequence, and given threshold, calculates the power swing in active power sequence sliding window
Value, and judge whether the power swing value is greater than the given threshold, wherein judge whether the power swing value is greater than institute
State given threshold, comprising:
If the power swing value is greater than the given threshold, the load signal of the real-time voltage electric current initial data is extracted
Transient state and stable state various dimensions feature are input to based on progress load identification in the trained classifier of load characteristic java standard library
And decomposition, output category result；
If the power swing value is less than or equal to the given threshold, the real-time voltage of subscriber household inlet is resurveyed
Electric current initial data.
4. the method for non-intrusion type household electricity remained capacity as described in claim 1, which is characterized in that in output category knot
Before fruit, the method also includes:
If the classifier can recognize that target load, the opening time and end time of the target load are recorded, and is tired out
Count electric flux when target load operation, output display result；
If the classifier not can recognize that target load, the characteristic value for extracting target load stores and is uploaded to main website progress
Load characteristic java standard library is updated after identification.
5. a kind of identifying system, which is characterized in that including data sampling unit, incident detection unit, load characteristic library unit, bear
Lotus recognition unit, storage unit, load recognition result display unit and load identify main website, the incident detection unit, load
Feature library unit, the storage unit, load identification main website are electrically connected with the load recognition unit, the data
Sampling unit and time probe unit are electrically connected, and the load recognition result display unit is electrically connected；Wherein,
The data sampling unit for acquiring the real-time voltage and electric current initial data of subscriber household inlet, and will be obtained
The initial data of the voltage and current taken is sent to the incident detection unit；
The incident detection unit is used for received primary voltage current data to be carried out wavelet threshold-value filter processing
Filtered voltage and current data calculate power sequence, and whether the mutation value that power sequence is detected by sliding window detection method is more than institute
The threshold value of setting；
The load characteristic library unit is stored in load characteristic for carrying out feature extraction and training to resident's common load equipment
Java standard library；
The load recognition unit, the primary voltage current data for will receive are handled and are calculated, and algorithm institute is obtained
The multidimensional characteristic information of the remained capacity needed, and be directed into classifier and decomposed and identified；
The storage unit, for storage section primary voltage current sampling data, the multidimensional characteristic number after processing and calculating
According to load recognition result data；
The load recognition result display unit, is handled for the result to load recognition unit, obtains each electricity consumption
Starting time, shut-in time and the power consumption within the corresponding period of equipment, and classification is carried out to every electric appliances and is shown；
The load identifies main website, the update for identification and load characteristic java standard library to unknown load.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910653275.8A CN110514889A (en) | 2019-07-19 | 2019-07-19 | A kind of method and system of non-intrusion type household electricity remained capacity |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910653275.8A CN110514889A (en) | 2019-07-19 | 2019-07-19 | A kind of method and system of non-intrusion type household electricity remained capacity |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110514889A true CN110514889A (en) | 2019-11-29 |
Family
ID=68622627
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910653275.8A Pending CN110514889A (en) | 2019-07-19 | 2019-07-19 | A kind of method and system of non-intrusion type household electricity remained capacity |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110514889A (en) |
Cited By (30)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110944413A (en) * | 2019-12-06 | 2020-03-31 | 江苏智臻能源科技有限公司 | Electric heat subdivision method based on historical load identification data under cloud edge cooperative architecture |
| CN111076368A (en) * | 2020-01-02 | 2020-04-28 | 珠海格力电器股份有限公司 | Air conditioner power consumption statistical method and system |
| CN111175599A (en) * | 2019-12-31 | 2020-05-19 | 广西电网有限责任公司电力科学研究院 | Identification method of non-intrusive air conditioner |
| CN111178393A (en) * | 2019-12-11 | 2020-05-19 | 广东浩迪智云技术有限公司 | Electric appliance power consumption classification metering method and device based on intelligent electric meter |
| CN111208375A (en) * | 2020-03-04 | 2020-05-29 | 威胜集团有限公司 | Real-time monitoring method and device for starting of fixed-frequency electric appliance and storage medium |
| CN111242391A (en) * | 2020-03-06 | 2020-06-05 | 云南电网有限责任公司电力科学研究院 | Machine learning model training method and system for power load identification |
| CN111308185A (en) * | 2020-02-28 | 2020-06-19 | 宁波三星医疗电气股份有限公司 | Non-invasive load identification method |
| CN111665407A (en) * | 2020-06-11 | 2020-09-15 | 国网山东省电力公司电力科学研究院 | Item electric quantity load identification system and method |
| CN111736012A (en) * | 2020-08-20 | 2020-10-02 | 国网浙江省电力有限公司湖州供电公司 | Non-invasive power load identification method and system |
| CN111751614A (en) * | 2020-06-29 | 2020-10-09 | 深圳供电局有限公司 | Non-invasive electricity consumption data testing system and method based on ModBus |
| CN111948446A (en) * | 2020-07-20 | 2020-11-17 | 江阴长仪集团有限公司 | Self-adaptive load identification method and intelligent electric energy meter |
| CN112016587A (en) * | 2020-07-09 | 2020-12-01 | 冯燕钧 | Energy consumption monitoring cloud collaborative non-invasive identification method based on master station feature library technology |
| CN112345824A (en) * | 2020-09-27 | 2021-02-09 | 国网湖南省电力有限公司 | Variable frequency air conditioner non-invasive identification method based on piecewise linear approximation algorithm |
| CN112560977A (en) * | 2020-12-23 | 2021-03-26 | 华北电力大学（保定） | Non-invasive load decomposition method based on sparse classifier hierarchical algorithm |
| CN112803463A (en) * | 2021-03-08 | 2021-05-14 | 浙江浙能电力股份有限公司萧山发电厂 | Electrochemical energy storage application scene evaluation method considering power grid operation constraint |
| CN112924790A (en) * | 2021-01-29 | 2021-06-08 | 上海逐鲨智能科技有限公司 | Fourier transform-based non-invasive load identification device and method thereof |
| CN113036754A (en) * | 2021-03-05 | 2021-06-25 | 重庆邮电大学 | Non-invasive load identification method based on improved DAG-SVMS |
| CN113191600A (en) * | 2021-04-13 | 2021-07-30 | 清科优能(深圳)技术有限公司 | Intelligent house non-invasive load identification method based on data mining |
| CN113238092A (en) * | 2021-05-25 | 2021-08-10 | 南京工程学院 | Non-invasive detection method based on machine learning |
| CN113255236A (en) * | 2021-07-07 | 2021-08-13 | 浙江大学 | Non-invasive load self-adaptive identification method based on twin network |
| CN113325248A (en) * | 2021-04-13 | 2021-08-31 | 清科优能(深圳)技术有限公司 | Intelligent house non-invasive load identification system based on edge calculation |
| CN113567722A (en) * | 2021-07-08 | 2021-10-29 | 浙江万胜智能科技股份有限公司 | Power control method and device for electrical appliance |
| CN114113773A (en) * | 2021-11-11 | 2022-03-01 | 国网辽宁省电力有限公司经济技术研究院 | Non-invasive load monitoring method based on zero sample learning |
| WO2022042070A1 (en) * | 2020-08-27 | 2022-03-03 | 广东工业大学 | Non-intrusive load monitoring method |
| CN114236234A (en) * | 2021-11-17 | 2022-03-25 | 国网天津市电力公司 | Electrical appliance characteristic identification method based on fundamental wave and harmonic wave mixed criterion |
| WO2022068534A1 (en) * | 2020-10-30 | 2022-04-07 | 青岛海尔洗衣机有限公司 | Control method for smart home system |
| CN114325081A (en) * | 2021-12-29 | 2022-04-12 | 润建股份有限公司 | Non-invasive load identification method based on multi-modal characteristics |
| CN115167273A (en) * | 2022-09-05 | 2022-10-11 | 中建安装集团有限公司 | Construction power consumption energy-saving management control system on construction site |
| CN115792472A (en) * | 2023-01-29 | 2023-03-14 | 北京志翔科技股份有限公司 | High-frequency data recording socket for load identification and load data identification method |
| CN116859169A (en) * | 2023-09-01 | 2023-10-10 | 国网浙江省电力有限公司余姚市供电公司 | Non-invasive power load decomposition method and system |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105186693A (en) * | 2015-09-28 | 2015-12-23 | 南方电网科学研究院有限责任公司 | Non-intrusive mode electrical load identification system and method |
| CN106443244A (en) * | 2016-08-31 | 2017-02-22 | 南方电网科学研究院有限责任公司 | Method and system for identifying type of electric devices |
| CN109145949A (en) * | 2018-07-19 | 2019-01-04 | 山东师范大学 | Non-intrusive electrical load monitoring and decomposition method and system based on integrated study |
| CN109492667A (en) * | 2018-10-08 | 2019-03-19 | 国网天津市电力公司电力科学研究院 | A kind of feature selecting discrimination method for non-intrusive electrical load monitoring |
| CN109813978A (en) * | 2018-12-25 | 2019-05-28 | 武汉中原电子信息有限公司 | A kind of non-intruding load-type recognition methods of variation characteristic between comprehensive transient characteristic and stable state |
| CN109870600A (en) * | 2019-04-04 | 2019-06-11 | 江苏林洋能源股份有限公司 | A kind of non-intrusion type load discrimination method and intelligent electric meter with self-learning capability |
| US20190195929A1 (en) * | 2007-09-18 | 2019-06-27 | Georgia Tech Research Corporation | Detecting actuation of electrical devices using electrical noise over a power line |
-
2019
- 2019-07-19 CN CN201910653275.8A patent/CN110514889A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20190195929A1 (en) * | 2007-09-18 | 2019-06-27 | Georgia Tech Research Corporation | Detecting actuation of electrical devices using electrical noise over a power line |
| CN105186693A (en) * | 2015-09-28 | 2015-12-23 | 南方电网科学研究院有限责任公司 | Non-intrusive mode electrical load identification system and method |
| CN106443244A (en) * | 2016-08-31 | 2017-02-22 | 南方电网科学研究院有限责任公司 | Method and system for identifying type of electric devices |
| CN109145949A (en) * | 2018-07-19 | 2019-01-04 | 山东师范大学 | Non-intrusive electrical load monitoring and decomposition method and system based on integrated study |
| CN109492667A (en) * | 2018-10-08 | 2019-03-19 | 国网天津市电力公司电力科学研究院 | A kind of feature selecting discrimination method for non-intrusive electrical load monitoring |
| CN109813978A (en) * | 2018-12-25 | 2019-05-28 | 武汉中原电子信息有限公司 | A kind of non-intruding load-type recognition methods of variation characteristic between comprehensive transient characteristic and stable state |
| CN109870600A (en) * | 2019-04-04 | 2019-06-11 | 江苏林洋能源股份有限公司 | A kind of non-intrusion type load discrimination method and intelligent electric meter with self-learning capability |
Non-Patent Citations (1)
| Title |
|---|
| 陈艳燕等: "《单分类支持张量机模型及算法研究》", 30 September 2017 * |
Cited By (43)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110944413A (en) * | 2019-12-06 | 2020-03-31 | 江苏智臻能源科技有限公司 | Electric heat subdivision method based on historical load identification data under cloud edge cooperative architecture |
| CN110944413B (en) * | 2019-12-06 | 2021-08-10 | 江苏智臻能源科技有限公司 | Electric heat subdivision method based on historical load identification data under cloud edge cooperative architecture |
| CN111178393A (en) * | 2019-12-11 | 2020-05-19 | 广东浩迪智云技术有限公司 | Electric appliance power consumption classification metering method and device based on intelligent electric meter |
| CN111175599A (en) * | 2019-12-31 | 2020-05-19 | 广西电网有限责任公司电力科学研究院 | Identification method of non-intrusive air conditioner |
| CN111175599B (en) * | 2019-12-31 | 2022-04-19 | 广西电网有限责任公司电力科学研究院 | Identification method of non-intrusive air conditioner |
| CN111076368A (en) * | 2020-01-02 | 2020-04-28 | 珠海格力电器股份有限公司 | Air conditioner power consumption statistical method and system |
| CN111076368B (en) * | 2020-01-02 | 2020-11-24 | 珠海格力电器股份有限公司 | Air conditioner power consumption statistical method and system |
| CN111308185A (en) * | 2020-02-28 | 2020-06-19 | 宁波三星医疗电气股份有限公司 | Non-invasive load identification method |
| CN111208375A (en) * | 2020-03-04 | 2020-05-29 | 威胜集团有限公司 | Real-time monitoring method and device for starting of fixed-frequency electric appliance and storage medium |
| CN111242391A (en) * | 2020-03-06 | 2020-06-05 | 云南电网有限责任公司电力科学研究院 | Machine learning model training method and system for power load identification |
| CN111665407A (en) * | 2020-06-11 | 2020-09-15 | 国网山东省电力公司电力科学研究院 | Item electric quantity load identification system and method |
| CN111751614B (en) * | 2020-06-29 | 2023-08-04 | 深圳供电局有限公司 | ModBus-based non-invasive electricity consumption data testing system and method thereof |
| CN111751614A (en) * | 2020-06-29 | 2020-10-09 | 深圳供电局有限公司 | Non-invasive electricity consumption data testing system and method based on ModBus |
| CN112016587A (en) * | 2020-07-09 | 2020-12-01 | 冯燕钧 | Energy consumption monitoring cloud collaborative non-invasive identification method based on master station feature library technology |
| CN111948446A (en) * | 2020-07-20 | 2020-11-17 | 江阴长仪集团有限公司 | Self-adaptive load identification method and intelligent electric energy meter |
| CN111736012A (en) * | 2020-08-20 | 2020-10-02 | 国网浙江省电力有限公司湖州供电公司 | Non-invasive power load identification method and system |
| CN111736012B (en) * | 2020-08-20 | 2021-01-05 | 国网浙江省电力有限公司湖州供电公司 | Non-invasive power load identification method and system |
| WO2022042070A1 (en) * | 2020-08-27 | 2022-03-03 | 广东工业大学 | Non-intrusive load monitoring method |
| GB2606284A (en) * | 2020-08-27 | 2022-11-02 | Univ Guangdong Technology | Non-intrusive load monitoring method |
| CN112345824B (en) * | 2020-09-27 | 2022-04-26 | 国网湖南省电力有限公司 | Variable frequency air conditioner non-invasive identification method based on piecewise linear approximation algorithm |
| CN112345824A (en) * | 2020-09-27 | 2021-02-09 | 国网湖南省电力有限公司 | Variable frequency air conditioner non-invasive identification method based on piecewise linear approximation algorithm |
| WO2022068534A1 (en) * | 2020-10-30 | 2022-04-07 | 青岛海尔洗衣机有限公司 | Control method for smart home system |
| CN112560977A (en) * | 2020-12-23 | 2021-03-26 | 华北电力大学（保定） | Non-invasive load decomposition method based on sparse classifier hierarchical algorithm |
| CN112560977B (en) * | 2020-12-23 | 2022-09-27 | 华北电力大学（保定） | Non-invasive load decomposition method based on sparse classifier hierarchical algorithm |
| CN112924790A (en) * | 2021-01-29 | 2021-06-08 | 上海逐鲨智能科技有限公司 | Fourier transform-based non-invasive load identification device and method thereof |
| CN113036754A (en) * | 2021-03-05 | 2021-06-25 | 重庆邮电大学 | Non-invasive load identification method based on improved DAG-SVMS |
| CN112803463A (en) * | 2021-03-08 | 2021-05-14 | 浙江浙能电力股份有限公司萧山发电厂 | Electrochemical energy storage application scene evaluation method considering power grid operation constraint |
| CN113325248A (en) * | 2021-04-13 | 2021-08-31 | 清科优能(深圳)技术有限公司 | Intelligent house non-invasive load identification system based on edge calculation |
| CN113191600A (en) * | 2021-04-13 | 2021-07-30 | 清科优能(深圳)技术有限公司 | Intelligent house non-invasive load identification method based on data mining |
| CN113238092A (en) * | 2021-05-25 | 2021-08-10 | 南京工程学院 | Non-invasive detection method based on machine learning |
| CN113238092B (en) * | 2021-05-25 | 2022-07-01 | 南京工程学院 | Non-invasive detection method based on machine learning |
| CN113255236B (en) * | 2021-07-07 | 2021-10-08 | 浙江大学 | Non-invasive load self-adaptive identification method based on twin network |
| CN113255236A (en) * | 2021-07-07 | 2021-08-13 | 浙江大学 | Non-invasive load self-adaptive identification method based on twin network |
| CN113567722A (en) * | 2021-07-08 | 2021-10-29 | 浙江万胜智能科技股份有限公司 | Power control method and device for electrical appliance |
| CN113567722B (en) * | 2021-07-08 | 2023-05-26 | 浙江万胜智能科技股份有限公司 | Power control method and device for electric appliance |
| CN114113773A (en) * | 2021-11-11 | 2022-03-01 | 国网辽宁省电力有限公司经济技术研究院 | Non-invasive load monitoring method based on zero sample learning |
| CN114113773B (en) * | 2021-11-11 | 2024-03-19 | 国网辽宁省电力有限公司经济技术研究院 | Non-invasive load monitoring method based on zero sample learning |
| CN114236234A (en) * | 2021-11-17 | 2022-03-25 | 国网天津市电力公司 | Electrical appliance characteristic identification method based on fundamental wave and harmonic wave mixed criterion |
| CN114325081A (en) * | 2021-12-29 | 2022-04-12 | 润建股份有限公司 | Non-invasive load identification method based on multi-modal characteristics |
| CN115167273A (en) * | 2022-09-05 | 2022-10-11 | 中建安装集团有限公司 | Construction power consumption energy-saving management control system on construction site |
| CN115792472A (en) * | 2023-01-29 | 2023-03-14 | 北京志翔科技股份有限公司 | High-frequency data recording socket for load identification and load data identification method |
| CN116859169A (en) * | 2023-09-01 | 2023-10-10 | 国网浙江省电力有限公司余姚市供电公司 | Non-invasive power load decomposition method and system |
| CN116859169B (en) * | 2023-09-01 | 2023-11-17 | 国网浙江省电力有限公司余姚市供电公司 | Non-invasive power load decomposition method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110514889A (en) | A kind of method and system of non-intrusion type household electricity remained capacity | |
| Ahmadi et al. | Load decomposition at smart meters level using eigenloads approach | |
| Jazizadeh et al. | An unsupervised hierarchical clustering based heuristic algorithm for facilitated training of electricity consumption disaggregation systems | |
| Lin et al. | Non-intrusive load monitoring by novel neuro-fuzzy classification considering uncertainties | |
| CN109975673B (en) | Method for identifying fault arc at direct current side of photovoltaic microgrid | |
| CN109470957B (en) | Non-invasive electrical appliance identification method based on Frey's distance | |
| CN109449919A (en) | A kind of non-intrusion type analysis of power consumption load method and device | |
| Li et al. | An enhanced ISODATA algorithm for recognizing multiple electric appliances from the aggregated power consumption dataset | |
| Mittelsdorf et al. | Submeter based training of multi-class support vector machines for appliance recognition in home electricity consumption data | |
| Yin et al. | Research on non-intrusive unknown load identification technology based on deep learning | |
| CN112819649A (en) | Method and device for determining station area subscriber change relationship | |
| CN109460419A (en) | A kind of equipment state change events monitoring method | |
| Saha et al. | Comprehensive NILM framework: device type classification and device activity status monitoring using capsule network | |
| CN109946543A (en) | V-I figure sample data set screening technique for non-intrusion type load monitoring | |
| CN113193654A (en) | Event-driven non-intrusive power load monitoring method based on transient and steady state combination characteristics | |
| Zhibin et al. | Research on household appliances recognition method based on data screening of deep learning | |
| Zhou et al. | Research on non-invasive load monitoring based on convolutional neural network | |
| Lin et al. | Application of neuro-fuzzy pattern recognition for non-intrusive appliance load monitoring in electricity energy conservation | |
| CN109616997A (en) | The optimization method and device of air switch, circuit | |
| CN115758188A (en) | Non-invasive load identification method, device, equipment and medium | |
| CN115563583A (en) | Non-invasive load monitoring method based on multi-physical quantity fusion | |
| Li et al. | A training-free non-intrusive load monitoring approach for high-frequency measurements based on graph signal processing | |
| CN114662576A (en) | Non-invasive storage battery car charging detection method and system based on supervised classification | |
| Djordjevic et al. | Nonintrusive identification and type recognition of household appliances based on the harmonic analysis of the steady-state current | |
| CN112949563A (en) | Non-invasive load identification method based on variable point detection and improved KNN algorithm |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191129 |