CN103018611A - Non-invasive load monitoring method and system based on current decomposition - Google Patents
Non-invasive load monitoring method and system based on current decomposition Download PDFInfo
- Publication number
- CN103018611A CN103018611A CN2012105795631A CN201210579563A CN103018611A CN 103018611 A CN103018611 A CN 103018611A CN 2012105795631 A CN2012105795631 A CN 2012105795631A CN 201210579563 A CN201210579563 A CN 201210579563A CN 103018611 A CN103018611 A CN 103018611A
- Authority
- CN
- China
- Prior art keywords
- consumer
- current
- vector
- module
- load monitoring
- 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.)
- Granted
Links
Images
Abstract
The invention provides non-invasive load monitoring method and system based on current decomposition. The method includes the steps of firstly, training each electrical device to obtain statistic manifolds of the electrical device respectively, and storing the statistic manifolds into a database; secondly, installing an acquisition unit at an inlet of a power grid, acquiring electrical data of all the electrical devices in real time, and converting the electrical data into vector representation of calculating statistic manifolds to acquire a vector set of total current; and thirdly, calculating statistic manifolds of the unit loading in the electrical devices respectively, and decomposing a total statistic manifold according to the statistic manifolds to acquire real-time current vectors of the electrical devices respectively. Acquired total current is directly decomposed to each device, detailed electrical information on related devices is calculated by current information directly instead of evaluating complex operation status, and accordingly accurate and detailed electrical information can be acquired.
Description
Technical field
The present invention relates to the Computer Applied Technology field, particularly a kind of non-intrusion type load monitoring method and system based on Current Decomposition.
Background technology
IBM in 2008 propose the concept of " the wisdom earth ", the wisdom earth are described as " more thorough perception, more fully interconnect, more deep intellectuality ".U.S. government proposes in the near future also intelligent grid to be brought into schedule as the pith in its " plan of new forms of energy rescue market " at the wisdom earth.Traditional electrical network of knowing with us by comparison, intelligent grid means and obtains as far as possible more information, more pays attention to mutual between the user, and the service that more puts in place is provided by resolving information.The ustomer premises access equipments such as intelligent electric meter directly contact with the user, and guides user electricity consumption behavior is the important embodiment that intelligent grid is different from traditional electrical network.By providing the detailed electricity consumption situation on the relevant devices, can effectively reduce the user to the understanding deviation of electricity consumption behavior, the use habit of optimizing user, thus obtain better power savings.
Therefore, effectively obtaining the detailed power information on the relevant devices in the power utilization environment (family, production environment etc.), is the gordian technique that intelligent grid field user client information gathers.In the situation that does not affect power utilization environment, obtain the monitoring technology of the detailed power information on each equipment from the outside, be called as non-intrusion type load monitoring (NILM) technology.Up to the present, non-intrusion type load monitoring technology mainly comprises two large classes:
1. based on the load monitoring technology of steady-state analysis: this type of technology at first defines a plurality of running statuses (stable state) for each consumer, and sets up characteristic of correspondence in the training stage for each running status; After beginning monitoring, by comparing of the global information that will collect and known characteristic set, obtain the running status of all consumers under the current state; Finally according to predefined running status, provide the detailed power information on the relevant devices.
2. based on the load monitoring technology of transient event: this type of technology at first defines a plurality of transient events for each consumer, and sets up characteristic of correspondence in the training stage for each transient event; After beginning monitoring, by comparing of the global information that will collect and known characteristic set, judge whether transient event occurs under the current state; When having transient event to occur, according to event definition, revise the running status of corresponding device, finally provide the detailed power information on the relevant devices.
Although above-mentioned two kinds of technology are all supported the non-intrusion type load monitoring, satisfy the demand that intelligent grid field user client information gathers, but these two kinds of technology have all been made similar hypothesis to consumer: equipment has metastable running status, determining can be according to known information after the running status, the detailed power information of equipment.And along with the epoch are progressive, day by day elastification of the behavior of consumer so that this hypothesis is no longer applicable, can not obtain accurately power information.For example: in the same time period, the computer of running game program is with respect to the more electric power of simple browsing page consumption, and idle and busy power consumption difference may surpass 30%.
Summary of the invention
For the elasticity electricity consumption behavior of consumer, this method directly decomposes the total current that collects on each equipment the phase space modeling of the electric current by equipment.No longer the running status of complexity is estimated, directly utilized the detailed power information on the current information calculating relevant devices.
For achieving the above object, the invention provides a kind of non-intrusion type load monitoring method based on Current Decomposition, the method comprises:
Step 1 is trained each consumer, obtains the statistical manifold of described each consumer, and deposits described statistical manifold in database;
Step 2 is installed collecting unit in the electrical network porch, the electricity consumption data of all consumers of Real-time Obtaining, and with the vector representation that described electricity consumption data-switching becomes counting statistics to flow shape, obtain the vector set of total current;
Step 3, computing unit are written into the statistical manifold of described each consumer, decompose described total statistical manifold according to described statistical manifold, thereby obtain the real-time current vector of each consumer.
Further, described step 1 comprises:
Step 11, electric current and voltage signal when gathering each consumer normal operation;
Step 12 with described electric current and voltage signal vectorization, obtains the vector set of each consumer;
Step 13 according to described vector set, constructs the statistical manifold of this consumer, until finish the training of all consumers.
Further, described step 2 comprises:
Step 21, electric current and the voltage signal of described all consumers of collecting unit Real-time Obtaining;
Step 22, described computing unit obtain the vector set of total current with described electric current and voltage signal vectorization.
Further, described step 3 comprises:
Step 31, described computing unit is written into the statistical manifold of each consumer, chooses the optimization algorithm of using when decomposing according to the character of statistical manifold;
Step 32 is used described optimization algorithm, in conjunction with the statistical manifold of described each consumer, the vector set of described total current is decomposed, and obtains the electric current on each equipment.
Described non-intrusion type load monitoring method also comprises:
Step 4, finish decomposition after, decomposition result is carried out error analysis;
Step 5, if error in accepting scope, described computing unit is according to the power of each consumer of real-time current vector calculation of described each consumer, and imports described power into display unit; Otherwise carry out abnormality processing.
For achieving the above object, the present invention also provides a kind of non-intrusion type load monitoring system based on Current Decomposition, and this system comprises:
Training module is trained each consumer, obtains the statistical manifold of described each consumer, and deposits described statistical manifold in database;
Pretreatment module is installed collecting unit in the electrical network porch, the electricity consumption data of all consumers of Real-time Obtaining, and described electricity consumption data-switching become the total current vector;
Decomposing module, computing unit are written into the statistical manifold of described each consumer, decompose described total current vector according to described statistical manifold, thereby obtain the real-time current vector of each consumer.
Further, described training module comprises:
The first acquisition module, electric current and voltage signal when gathering each consumer normal operation;
The primary vector module with described electric current and voltage signal vectorization, obtains the vector set of each consumer;
First-class shape constructing module according to described vector set, constructs the statistical manifold of this consumer, until finish the training of all consumers.
Further, described pretreatment module comprises:
The second acquisition module, electric current and the voltage signal of described all consumers of collecting unit Real-time Obtaining;
Secondary vector module, described computing unit obtain the vector set of all consumers with described electric current and voltage signal vectorization.
Further, described decomposing module comprises:
The algorithm picks module, described computing unit is written into the statistical manifold of each consumer, chooses the optimization algorithm of using when decomposing according to the character of statistical manifold;
The algorithm execution module uses described optimization algorithm, in conjunction with the statistical manifold of described each consumer, the vector set of described total current is decomposed, and obtains the electric current on each equipment.
Described non-intrusion type load monitoring system also comprises:
Analysis module, finish decomposition after, decomposition result is carried out error analysis;
Processing module, if error in accepting scope, described computing unit is according to the power of each consumer of real-time current vector calculation of described each consumer, and imports described power into display unit; Otherwise carry out abnormality processing.
Beneficial functional of the present invention is,
1. the present invention does not rely on the judgement of equipment state, directly decomposes electric current: existing non-intrusion type load monitoring technology is at first carried out the estimation of equipment state mostly, finishes afterwards the analysis of power information again.These class methods have a direct defective, and when having the elastic devices of load dynamic change in the family, analysis precision can significantly decrease.And the present invention does not rely on the running status judgement of equipment directly to the electric current phase space modeling of equipment, can effectively decompose the electric current of the elastic devices generation of multiple load dynamic change.
2. the present invention is in the computing method of decomposing electric current, use multiple optimization method, greatly improve the Current Decomposition precision: existing non-intrusion type load monitoring technology is at first carried out the estimation of equipment state mostly, the estimation accuracy of this step is greatly about about 95%, the power information analysis precision that carries out on this basis is greatly about about 90%, when having a plurality of elastic devices in the family, precision also has the downslide about 10%.And the present invention has simplified calculation procedure, uses simultaneously the direct resolving device electric current of optimization method, and the curent change situation of effectively following the trail of each equipment can bring up to 95% with the mean accuracy of Current Decomposition.
Describe the present invention below in conjunction with the drawings and specific embodiments, but not as a limitation of the invention.
Description of drawings
Fig. 1 is the non-intrusion type load monitoring method flow diagram based on Current Decomposition of the present invention;
Fig. 2 is the non-intrusion type load monitoring system schematic based on Current Decomposition of the present invention;
Fig. 3 is training stage process flow diagram of the present invention;
Fig. 4 is catabolic phase process flow diagram of the present invention;
Fig. 5 is consumer structural representation of the present invention.
Embodiment
Fig. 5 is consumer structural representation of the present invention.This structure realizes comprising three critical pieces: collecting unit, computing unit and display unit, wherein:
Collecting unit is main input media, and consumer Current Decomposition system obtains the current-voltage information in the whole family by near the domestic electric network entrance collecting unit being installed.Current sensor in the collecting unit can use the magnetic field induction chip based on Hall effect, utilizes the current data collection of the principle realization non-intrusion type of electromagnetic induction.And voltage sensor directly access the mains circuit can be in parallel with all consumers, measure the voltage on it.In decomposition algorithm, when calculating the vector representation of signal, require the data acquisition frequency to be higher than a certain threshold value, in one embodiment of the invention: when algorithm uses Short Time Fourier Transform, require to comprise in each cycle 256 sampled points, the sample frequency of 12.8kHz namely need to be arranged the alternating current of 50Hz.
Computing unit is the core of whole decomposing system, and the current-voltage information of whole family is broken down on each consumer herein, obtains the real-time power information of each equipment.Computing unit is in the central authorities of whole decomposing system, obtains the data of Real-time Collection from collecting unit, after calculating decomposition result is delivered to display unit.In an embodiment of the present invention, possessing arbitrarily the equipment of enough computing powers can be as computing unit in the system.Therefore, computing unit both can independently be realized, also can share a processor with collecting unit, consisted of the novel collecting device of similar intelligent electric meter; Can also share a processor with display unit, consist of the novel display device of similar intelligent terminal.
Display unit is the formant of system and user interactions, except basic demonstration and interactive function, also needs and can add up and analyze real-time power information, and the storage and management function of supported data.Display unit gets access to the real-time power information of each consumer in the family from computing unit, historical power information and real-time power information to each consumer and whole family are upgraded, and these contents are saved in database or other media; On the other hand, display unit need to be realized the effective User Interface of a cover, comprises patterned display interface, and the form with power information can be understood to be converted into the user feeds back to the user; The user can point out the part of systematic analysis distortion simultaneously, and feeds back to decomposing system, thus the precision of decomposing after improving.
Gordian technique of the present invention concentrates in the employed Current Decomposition algorithm of computing unit.Algorithm flow can be divided into two stages: training stage and catabolic phase.
Training stage is the necessary stage of setting up computation model for each consumer.Electric current and voltage signal when in this one-phase, gathering respectively each equipment normal operation; Afterwards the data that collect are converted to predefined vector representation; Statistical manifold by the real data structure current device electric current phase space that collects.This stage does not need to carry out simultaneously the training of all electrical equipment, therefore can be finished in the family by user oneself, can have equipment supplier or the third-party institution to finish yet.But from the effect angle of training and realize that required labor capacity of this stage analyzes, finish the training stage by equipment supplier or the third-party institution, and the result is sent to each user, can be more accurately and effectively finish this stage.Each statistical manifold that obtains in this stage will be stored into database, use in next stage.
Catabolic phase is the Main Stage of carrying out the consumer Current Decomposition.In this one-phase, need collecting unit is installed on the electrical network porch, gather electric current and the voltage signal of whole family; Then import the data that collect into computing unit, be converted to vector representation; From database, take out simultaneously the statistical manifold of all known devices in the family, and choose suitable optimization method according to the mathematical feature of each statistical manifold; In the situation of known total power information, by calculating in the whole family current distributions most possible on each equipment; Current information by each equipment that obtains is derived concrete power information, and these communications are finished to display unit and shown and follow-up statistical treatment the most at last.
Fig. 1 is the non-intrusion type load monitoring method flow diagram based on Current Decomposition of the present invention.As shown in Figure 1, the method comprises:
Step 1 is trained each consumer, obtains the statistical manifold of described each consumer, and deposits described statistical manifold in database;
Step 2 is installed collecting unit in the electrical network porch, the electricity consumption data of all consumers of Real-time Obtaining, and with the vector representation that described electricity consumption data-switching becomes counting statistics to flow shape, obtain the vector set of total current;
Step 3, computing unit are written into the statistical manifold of described each consumer, decompose described total statistical manifold according to described statistical manifold, thereby obtain the real-time current vector of each consumer.
The vector representation of settling signal.Utilize the methods such as Short Time Fourier Transform or wavelet transformation, the signals such as current/voltage that obtain are transformed into frequency domain from time domain, the signal waveform of a basic cycle is converted to a isolated point in the high dimension vector space, uses the high dimension vector of a frequency-region signal or the circuit signal waveform of matrix representation a period of time; Technique effect: after the conversion, signal more visual representation goes out physical significance on it, for example, can support to harmonic power the direct calculating of the physical quantitys such as reactive power.
The statistical models of apparatus for establishing electric current phase space.Utilize concept and the statistical correlation technique of stream shape in the infinitesimal geometry, power information according to the individual equipment that gathers in advance, the current signal phase space (signal possibility region in the high dimension vector space) of each equipment is expressed as a statistical manifold, it is the level and smooth probability distribution function of a series of support infinitesimal geometry methods, wherein, the geometry of stream shape is from the angle of Circuit theory equipment to be portrayed, and the value that the probability distribution on it has then been explained equipment electric current when operation distributes; Technique effect: to each consumer, all values of electric current are included in the statistical manifold on it, can utilize this stream shape to calculate corresponding device with the probability of specific currents work.
The computing method of structure Current Decomposition.According to the optimum solving method in the multiple operational research, in conjunction with statistical manifold corresponding to each known equipment, set up multiple Current Decomposition method; When needs decompose, can according to circumstances choose effective method, finish the decomposition to total current; Technique effect: for the household electricity environment of complexity, by this compound decomposition method, from the total current that gathers, decomposite the electric current by each equipment, and guarantee in system's normal operation, to decompose the average current precision of acquisition more than 95%.
Further, described step 1 comprises:
Step 11, electric current and voltage signal when gathering each consumer normal operation;
Step 12 with described electric current and voltage signal vectorization, obtains the vector set of each consumer;
Step 13 according to described vector set, constructs the statistical manifold of this consumer, until finish the training of all consumers.
Further, described step 2 comprises:
Step 21, electric current and the voltage signal of described all consumers of collecting unit Real-time Obtaining;
Step 22, described computing unit obtain the vector set of total current with described electric current and voltage signal vectorization.
Further, described step 3 comprises:
Step 31, described computing unit is written into the statistical manifold of each consumer, chooses the optimization algorithm of using when decomposing according to the character of statistical manifold.
Step 32 is used described optimization algorithm, in conjunction with the statistical manifold of described each consumer, the vector set of described total current is decomposed, and obtains the electric current on each equipment.
The optimization algorithm of mentioning in the described step 31 mainly comprises: least square method, maximal possibility estimation, minimum risk method and minimize maximum entropy method.
Described non-intrusion type load monitoring method also comprises:
Step 4, finish decomposition after, decomposition result is carried out error analysis;
Step 5, if error in accepting scope, described computing unit is according to the power of each consumer of real-time current vector calculation of described each consumer, and imports described power into display unit; Otherwise carry out abnormality processing.
Fig. 3 is training stage process flow diagram of the present invention.As shown in Figure 3, after the training stage begins, need to judge whether to have finished the training of all consumers.If not then choose a Devices to test, dispose training environment.The current-voltage information that training stage need to gather each consumer when normally moving, then computing unit carries out Short Time Fourier Transform to the current information that collects, thereby obtains each constantly vector representation corresponding to electric current in the frequency domain space.
Can use the current information in the 256 frequency domain vector indication equipment one-periods of tieing up, image data need guarantee each cycle sampled point more than 256, therefore selects 12.8kHz as sample frequency.And to specific consumer, after the current vector that obtains the sufficiently long time (all working state that so-called " sufficiently long time " may occur when comprising that this equipment normally uses in referring to during this period of time), begin afterwards to calculate the statistical manifold under its working current.
This method is with the mathematical abstractions of the equally distributed linear manifold of probability density as description device current phase space, computing method are principal component analysis (PCA): will collect all current vectors (row vector) and form a matrix, this matrix is carried out principal component analysis (PCA), choose variance contribution ratio greater than the vectorial linear manifold of constructing this equipment of 95% major component characteristic of correspondence, namely current vector must satisfy the linear manifold equation.Deposit the proper vector that obtains in database, use for catabolic phase.Until all consumers are finished training, the training stage finishes.
Fig. 4 is catabolic phase process flow diagram of the present invention.As shown in Figure 4, catabolic phase decomposes family's total current of Real-time Collection according to the electric current phase space of all devices of training stage acquisition.In catabolic phase, the first step is the statistical manifold that is written into all electrical equipment (consumer) the family from database.Open collecting unit, obtain real-time total current, voltage data.
Then whether judgement is received to decompose and cease and desist order at this moment, if do not receive, then utilize collecting unit to obtain total current data in the family in the electrical network porch, the data acquisition frequency is identical with the training stage, and computing unit obtains frequency domain vector corresponding to current time total current with data through Short Time Fourier Transform and represents.Then choose decomposition algorithm according to field condition, carry out Current Decomposition.Decomposition method in the present embodiment is chosen least square method,
By Kirchhoff's law as can be known, any time, total current is the electric current sum that equals on each equipment, and the current vector of each equipment all belongs to its linear manifold simultaneously.Therefore in catabolic phase, known total current and each electrical equipment linear manifold namely can obtain the system of equations that is comprised of two class equations: the equation that Kirchhoff's equation is corresponding with linear manifold.This system of equations is the overdetermined equation group, can not use common method for solving.Use least square method solving equation group in this decomposition stage, find the solution the electric current on each equipment of error minimize.In this decomposition stage, need consider contingent abnormal conditions in the solution procedure (using new equipment, device damage etc.).
After finishing decomposition computation, decomposition result is carried out error analysis: relatively decompose each current value sum obtain and the numerical relation of total current, by to the wherein analysis of deviate, judged whether unusual generation, and to unusually processing of may occurring.For normal decomposition result, import display unit into from computing unit and finish last data processing, according to the current vector on each equipment, in conjunction with the information of voltage in the family, calculate power and energy consumption on each equipment, and deposit the result in database and consult in order to the user.
In this decomposition stage, the current vector that obtains after the decomposition is the electric current frequency domain representation, can directly calculate power information with corresponding voltage vector constantly, need not to convert back time domain.In the present embodiment, catabolic phase will continue to carry out always, decompose the order that stops until the user sends decomposing system.
Fig. 2 is the non-intrusion type load monitoring system flowchart based on Current Decomposition of the present invention.As shown in Figure 2, this system comprises:
Decomposing module 300, computing unit are written into the statistical manifold of described each consumer, decompose described total statistical manifold according to described statistical manifold, thereby obtain the real-time current vector of each consumer.
The vector representation of settling signal.Utilize the methods such as Short Time Fourier Transform or wavelet transformation, the signals such as current/voltage that obtain are transformed into frequency domain from time domain, the signal waveform of a basic cycle is converted to a isolated point in the high dimension vector space, uses the high dimension vector of a frequency-region signal or the circuit signal waveform of matrix representation a period of time; Technique effect: after the conversion, signal more visual representation goes out physical significance on it, for example, can support to harmonic power the direct calculating of the physical quantitys such as reactive power.
The statistical models of apparatus for establishing electric current phase space.Utilize concept and the statistical correlation technique of stream shape in the infinitesimal geometry, power information according to the individual equipment that gathers in advance, the current signal phase space (signal possibility region in the high dimension vector space) of each equipment is expressed as a statistical manifold, it is the level and smooth probability distribution function of a series of support infinitesimal geometry methods, wherein, the geometry of stream shape is from the angle of Circuit theory equipment to be portrayed, and the value that the probability distribution on it has then been explained equipment electric current when operation distributes; Technique effect: to each consumer, all values of electric current are included in the statistical manifold on it, can utilize this stream shape to calculate corresponding device with the probability of specific currents work.
The computing method of structure Current Decomposition.According to the optimum solving method in the multiple operational research, in conjunction with statistical manifold corresponding to each known equipment, set up multiple Current Decomposition method; When needs decompose, can according to circumstances choose effective method, finish the decomposition to total current; Technique effect: for the household electricity environment of complexity, by this compound decomposition method, from the total current that gathers, decomposite the electric current by each equipment, and guarantee in system's normal operation, to decompose the average current precision of acquisition more than 95%.
Further, described training module 100 comprises:
The first acquisition module 110, electric current and voltage signal when gathering each consumer normal operation;
First-class shape constructing module 130 according to described vector set, constructs the statistical manifold of this consumer, until finish the training of all consumers.
Further, described pretreatment module 200 comprises:
The second acquisition module 210, electric current and the voltage signal of described all consumers of collecting unit Real-time Obtaining;
Further, described decomposing module 300 comprises:
Algorithm picks module 310, described computing unit is written into the statistical manifold of each consumer, chooses the optimization algorithm of using when decomposing according to the character of statistical manifold.
The optimization algorithm of mentioning in the described algorithm picks module 310 mainly comprises: least square method, maximal possibility estimation, minimum risk method and minimize maximum entropy method.
Described non-intrusion type load monitoring system also comprises:
Analysis module 400, finish decomposition after, decomposition result is carried out error analysis;
Processing module 500, if error in accepting scope, described computing unit is according to the power of each consumer of real-time current vector calculation of described each consumer, and imports described power into display unit; Otherwise carry out abnormality processing.
Fig. 3 is training stage process flow diagram of the present invention.As shown in Figure 3, after the training stage begins, need to judge whether to have finished the training of all consumers.If not then choose a Devices to test, dispose training environment.The current-voltage information that training stage need to gather each consumer when normally moving, then computing unit carries out Short Time Fourier Transform to the current information that collects, thereby obtains each constantly vector representation corresponding to electric current in the frequency domain space.
Can use the current information in the 256 frequency domain vector indication equipment one-periods of tieing up, image data need guarantee each cycle sampled point more than 256, therefore selects 12.8kHz as sample frequency.And to specific consumer, after the current vector that obtains the sufficiently long time (all working state that so-called " sufficiently long time " may occur when comprising that this equipment normally uses in referring to during this period of time), begin afterwards to calculate the statistical manifold under its working current.
This method is with the mathematical abstractions of the equally distributed linear manifold of probability density as description device current phase space, computing method are principal component analysis (PCA): will collect all current vectors (row vector) and form a matrix, this matrix is carried out principal component analysis (PCA), choose variance contribution ratio is constructed this equipment greater than 95% major component characteristic of correspondence vector linear manifold.Deposit the proper vector that obtains in database, use for catabolic phase.Until all consumers are finished training, the training stage finishes.
Fig. 4 is catabolic phase process flow diagram of the present invention.As shown in Figure 4, catabolic phase decomposes family's total current of Real-time Collection according to the electric current phase space of all devices of training stage acquisition.In catabolic phase, the first step is the statistical manifold that is written into all electrical equipment (consumer) the family from database.Open collecting unit, obtain real-time total current, voltage data.
Then whether judgement is received to decompose and cease and desist order at this moment, if do not receive, then utilize collecting unit to obtain total current data in the family in the electrical network porch, the data acquisition frequency is identical with the training stage, and computing unit obtains frequency domain vector corresponding to current time total current with data through Short Time Fourier Transform and represents.Then choose decomposition algorithm according to field condition, carry out Current Decomposition.By Kirchhoff's law as can be known, any time, total current is the electric current sum that equals on each equipment, and the current vector of each equipment all belongs to its linear manifold simultaneously.
Therefore in catabolic phase, known total current and each electrical equipment linear manifold namely can obtain the system of equations that is comprised of two class equations: the equation that Kirchhoff's equation is corresponding with linear manifold.Use least square method solving equation group in this decomposition stage, find the solution the electric current on each equipment of error minimize.In this decomposition stage, need consider contingent abnormal conditions in the solution procedure (using new equipment, device damage etc.).
After finishing decomposition computation, decomposition result is carried out error analysis, relatively decompose each current value obtain and the numerical relation of total current, judged whether unusual generation, and to unusually processing of may occurring.For normal decomposition result, import display unit into from computing unit and finish last data processing, according to the current vector on each equipment, in conjunction with the information of voltage in the family, calculate power and energy consumption on each equipment, and deposit the result in database and consult in order to the user.
In this decomposition stage, the current vector that obtains after the decomposition is the electric current frequency domain representation, can directly calculate power information with corresponding voltage vector constantly, need not to convert back time domain.In the present embodiment, catabolic phase will continue to carry out always, decompose the order that stops until the user sends decomposing system.
Certainly; the present invention also can have other various embodiments; in the situation that does not deviate from spirit of the present invention and essence thereof; those of ordinary skill in the art work as can make according to the present invention various corresponding changes and distortion, but these corresponding changes and distortion all should belong to the protection domain of the appended claim of the present invention.
Claims (10)
1. the non-intrusion type load monitoring method based on Current Decomposition is characterized in that, comprising:
Step 1 is trained each consumer, obtains the statistical manifold of described each consumer, and deposits described statistical manifold in database;
Step 2 is installed collecting unit in the electrical network porch, the electricity consumption data of all consumers of Real-time Obtaining, and with the vector representation that described electricity consumption data-switching becomes counting statistics to flow shape, obtain the vector set of total current;
Step 3, computing unit are written into the statistical manifold of described each consumer, decompose described total statistical manifold according to described statistical manifold, thereby obtain the real-time current vector of each consumer.
2. non-intrusion type load monitoring method as claimed in claim 1 is characterized in that described step 1 comprises:
Step 11, electric current and voltage signal when gathering each consumer normal operation;
Step 12 with described electric current and voltage signal vectorization, obtains the vector set of each consumer;
Step 13 according to described vector set, constructs the statistical manifold of this consumer, until finish the training of all consumers.
3. non-intrusion type load monitoring method as claimed in claim 1 is characterized in that described step 2 comprises:
Step 21, electric current and the voltage signal of described all consumers of collecting unit Real-time Obtaining;
Step 22, described computing unit obtain the vector set of total current with described electric current and voltage signal vectorization.
4. non-intrusion type load monitoring method as claimed in claim 1 is characterized in that described step 3 comprises:
Step 31, described computing unit is written into the statistical manifold of each consumer, chooses the optimization algorithm of using when decomposing according to the character of statistical manifold;
Step 32 is used described optimization algorithm, in conjunction with the statistical manifold of described each consumer, the vector set of described total current is decomposed, and obtains the electric current on each equipment.
5. non-intrusion type load monitoring method as claimed in claim 1 is characterized in that, described non-intrusion type load monitoring method also comprises:
Step 4, finish decomposition after, decomposition result is carried out error analysis;
Step 5, if error in accepting scope, described computing unit is according to the power of each consumer of real-time current vector calculation of described each consumer, and imports described power into display unit; Otherwise carry out abnormality processing.
6. the non-intrusion type load monitoring system based on Current Decomposition is characterized in that, comprising:
Training module is trained each consumer, obtains the statistical manifold of described each consumer, and deposits described statistical manifold in database;
Pretreatment module is installed collecting unit in the electrical network porch, the electricity consumption data of all consumers of Real-time Obtaining, and with the vector representation that described electricity consumption data-switching becomes counting statistics to flow shape, obtain the vector set of total current;
Decomposing module, computing unit are written into the statistical manifold of described each consumer, decompose described total statistical manifold according to described statistical manifold, thereby obtain the real-time current vector of each consumer.
7. non-intrusion type load monitoring as claimed in claim 6 system is characterized in that described training module comprises:
The first acquisition module, electric current and voltage signal when gathering each consumer normal operation;
The primary vector module with described electric current and voltage signal vectorization, obtains the vector set of each consumer;
First-class shape constructing module according to described vector set, constructs the statistical manifold of this consumer, until finish the training of all consumers.
8. non-intrusion type load monitoring as claimed in claim 6 system is characterized in that described pretreatment module comprises:
The second acquisition module, electric current and the voltage signal of described all consumers of collecting unit Real-time Obtaining;
Secondary vector module, described computing unit obtain the vector set of total current with described electric current and voltage signal vectorization.
9. non-intrusion type load monitoring as claimed in claim 6 system is characterized in that described decomposing module comprises:
The algorithm picks module, described computing unit is written into the statistical manifold of each consumer, chooses the optimization algorithm of using when decomposing according to the character of statistical manifold;
The algorithm execution module uses described optimization algorithm, in conjunction with the statistical manifold of described each consumer, the vector set of described total current is decomposed, and obtains the electric current on each equipment.
10. non-intrusion type load monitoring as claimed in claim 6 system is characterized in that, described non-intrusion type load monitoring system also comprises:
Analysis module, finish decomposition after, decomposition result is carried out error analysis;
Processing module, if error in accepting scope, described computing unit is according to the power of each consumer of real-time current vector calculation of described each consumer, and imports described power into display unit; Otherwise carry out abnormality processing.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210579563.1A CN103018611B (en) | 2012-12-27 | 2012-12-27 | Non-invasive load monitoring method and system based on current decomposition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210579563.1A CN103018611B (en) | 2012-12-27 | 2012-12-27 | Non-invasive load monitoring method and system based on current decomposition |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103018611A true CN103018611A (en) | 2013-04-03 |
| CN103018611B CN103018611B (en) | 2015-06-17 |
Family
ID=47967443
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210579563.1A Active CN103018611B (en) | 2012-12-27 | 2012-12-27 | Non-invasive load monitoring method and system based on current decomposition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103018611B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104483575A (en) * | 2014-12-22 | 2015-04-01 | 天津求实智源科技有限公司 | Self-adaptive load event detection method for noninvasive power monitoring |
| CN105652118A (en) * | 2015-12-29 | 2016-06-08 | 国家电网公司 | Load instantaneous energy feature-based power grid electric energy load monitoring method |
| CN106093630A (en) * | 2016-06-02 | 2016-11-09 | 华北电力大学 | A kind of non-intrusion type household electrical appliance discrimination method |
| CN106443233A (en) * | 2016-08-26 | 2017-02-22 | 北京电力经济技术研究院 | Non-invasive steady-state load monitoring method |
| CN110995543A (en) * | 2019-12-18 | 2020-04-10 | 云南大学 | Non-invasive method for monitoring abnormal internet surfing behavior of minors |
| CN112083287A (en) * | 2020-09-08 | 2020-12-15 | 上海交通大学 | Power distribution network dynamic event positioning method and system based on characteristic value spectrum distribution model |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO1997025625A1 (en) * | 1996-01-05 | 1997-07-17 | Massachusetts Institute Of Technology | Transient event detector for monitoring electrical loads |
| CN101567559A (en) * | 2009-06-04 | 2009-10-28 | 天津天大求实电力新技术股份有限公司 | Tabular method of non-intrusive electrical load decomposition |
| CN101576580A (en) * | 2009-06-04 | 2009-11-11 | 天津天大求实电力新技术股份有限公司 | Non-invasive unitized current on-line measurement method of electric equipment |
| WO2012082802A2 (en) * | 2010-12-13 | 2012-06-21 | Fraunhofer Usa, Inc. | Methods and system for nonintrusive load monitoring |
-
2012
- 2012-12-27 CN CN201210579563.1A patent/CN103018611B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO1997025625A1 (en) * | 1996-01-05 | 1997-07-17 | Massachusetts Institute Of Technology | Transient event detector for monitoring electrical loads |
| CN101567559A (en) * | 2009-06-04 | 2009-10-28 | 天津天大求实电力新技术股份有限公司 | Tabular method of non-intrusive electrical load decomposition |
| CN101576580A (en) * | 2009-06-04 | 2009-11-11 | 天津天大求实电力新技术股份有限公司 | Non-invasive unitized current on-line measurement method of electric equipment |
| WO2012082802A2 (en) * | 2010-12-13 | 2012-06-21 | Fraunhofer Usa, Inc. | Methods and system for nonintrusive load monitoring |
Non-Patent Citations (1)
| Title |
|---|
| 黎鹏: "非侵入式电力负荷分解与监测", 《万方数据企业知识服务平台》, 30 August 2010 (2010-08-30) * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104483575A (en) * | 2014-12-22 | 2015-04-01 | 天津求实智源科技有限公司 | Self-adaptive load event detection method for noninvasive power monitoring |
| CN104483575B (en) * | 2014-12-22 | 2017-05-03 | 天津求实智源科技有限公司 | Self-adaptive load event detection method for noninvasive power monitoring |
| CN105652118A (en) * | 2015-12-29 | 2016-06-08 | 国家电网公司 | Load instantaneous energy feature-based power grid electric energy load monitoring method |
| CN105652118B (en) * | 2015-12-29 | 2019-02-26 | 国家电网公司 | A kind of power grid power budget monitoring method based on load instantaneous energy feature |
| CN106093630A (en) * | 2016-06-02 | 2016-11-09 | 华北电力大学 | A kind of non-intrusion type household electrical appliance discrimination method |
| CN106093630B (en) * | 2016-06-02 | 2019-01-15 | 华北电力大学 | A kind of non-intrusion type household electrical appliance discrimination method |
| CN106443233A (en) * | 2016-08-26 | 2017-02-22 | 北京电力经济技术研究院 | Non-invasive steady-state load monitoring method |
| CN110995543A (en) * | 2019-12-18 | 2020-04-10 | 云南大学 | Non-invasive method for monitoring abnormal internet surfing behavior of minors |
| CN110995543B (en) * | 2019-12-18 | 2022-11-01 | 云南大学 | Non-invasive method for monitoring abnormal internet surfing behavior of minors |
| CN112083287A (en) * | 2020-09-08 | 2020-12-15 | 上海交通大学 | Power distribution network dynamic event positioning method and system based on characteristic value spectrum distribution model |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103018611B (en) | 2015-06-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| de Souza et al. | Data compression in smart distribution systems via singular value decomposition | |
| Mohammed et al. | Available transfer capability calculation methods: A comprehensive review | |
| CN103135009B (en) | Electric appliance detection method and system based on user feedback information | |
| CN109345409B (en) | Comprehensive energy efficiency management method for residential users based on broadband carrier | |
| CN103018611B (en) | Non-invasive load monitoring method and system based on current decomposition | |
| CN107294120B (en) | Active power distribution network hybrid energy storage capacity optimal configuration method and device | |
| CN103872782A (en) | Electric energy quality data comprehensive service system | |
| CN103001230A (en) | Non-invasive power load monitoring and decomposing current mode matching method | |
| Al-Ghaili et al. | Energy management systems and strategies in buildings sector: A scoping review | |
| CN112364516A (en) | 10kV feeder line optimal load capacity calculation method considering different load structures | |
| Adinolfi et al. | Pseudo-measurements modeling using neural network and Fourier decomposition for distribution state estimation | |
| Zang et al. | A robust state estimator for integrated electrical and heating networks | |
| CN113036786A (en) | Low-voltage distribution transformer user phase sequence identification and three-phase imbalance adjustment method | |
| CN115759708A (en) | Line loss analysis method and system considering power space-time distribution characteristics | |
| CN108400584A (en) | A kind of micro-capacitance sensor method for diagnosing faults based on correlation analysis matching degree | |
| CN107732902B (en) | Power distribution network economic operation monitoring and evaluation method | |
| Miraftabzadeh et al. | K-means and alternative clustering methods in modern power systems | |
| CN112182499A (en) | Low-voltage distribution network topological structure identification method based on time sequence electric quantity data | |
| Faranda et al. | SCADA system for optimization of energy exchange with the BESS in a residential case | |
| Domínguez et al. | Analysis of electric power consumption using Self-Organizing Maps. | |
| Yang et al. | Probabilistic revenue analysis of microgrid considering source-load and forecast uncertainties | |
| Cheung et al. | Towards improved real-time observability of behind-meter photovoltaic systems: A data-driven approach | |
| Hu et al. | Transformer-customer relationship identification for low-voltage distribution networks based on joint optimization of voltage silhouette coefficient and power loss coefficient | |
| Deng et al. | Medium-term rolling load forecasting based on seasonal decomposition and long short-term memory neural network | |
| Athanasiadis et al. | A Benchmarking Testbed for Low-Voltage Active Distribution Network Studies |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |