CN113269227A - Non-invasive deep decomposition method and system for small and miniature load time-space electricity consumption behavior - Google Patents
Non-invasive deep decomposition method and system for small and miniature load time-space electricity consumption behavior Download PDFInfo
- Publication number
- CN113269227A CN113269227A CN202110405550.1A CN202110405550A CN113269227A CN 113269227 A CN113269227 A CN 113269227A CN 202110405550 A CN202110405550 A CN 202110405550A CN 113269227 A CN113269227 A CN 113269227A
- Authority
- CN
- China
- Prior art keywords
- time
- event
- electric appliance
- electricity consumption
- consumption 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
- 230000005611 electricity Effects 0.000 title claims abstract description 84
- 238000000354 decomposition reaction Methods 0.000 title claims abstract description 28
- 238000000034 method Methods 0.000 title claims abstract description 21
- 230000006399 behavior Effects 0.000 claims abstract description 37
- 239000011159 matrix material Substances 0.000 claims description 24
- 238000007405 data analysis Methods 0.000 claims description 12
- 238000007621 cluster analysis Methods 0.000 claims description 11
- 238000001514 detection method Methods 0.000 claims description 11
- 238000000605 extraction Methods 0.000 claims description 9
- 238000007476 Maximum Likelihood Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 2
- 238000013480 data collection Methods 0.000 claims 2
- 239000000126 substance Substances 0.000 claims 1
- 230000000694 effects Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 238000011144 upstream manufacturing Methods 0.000 description 2
- 240000007594 Oryza sativa Species 0.000 description 1
- 235000007164 Oryza sativa Nutrition 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 235000009566 rice Nutrition 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
Abstract
The invention discloses a non-invasive deep decomposition method and a non-invasive deep decomposition system for space-time electric behaviors of small and miniature loads. The identification of the power utilization behaviors of the small micro loads on the time scale and the space scale has important application value in the field of intelligent power utilization of a power system, but at present, the identification of the power utilization behaviors of the small micro loads on the time scale and the space scale is difficult through electric appliance data. The invention is based on the intelligent circuit breaker which can simultaneously collect the electricity utilization data of the main circuit and each branch circuit, and the electricity utilization data is mined by the steps of collecting the electricity utilization data of the main circuit and different branch circuits, detecting an electric appliance event, extracting time-space characteristics of the event, matching the electric appliance and the like, so that the electricity utilization behaviors of the small micro load in time scale and space scale are identified and obtained, and the time-space electricity utilization behaviors of the electric appliance are deeply sensed in real time.
Description
Technical Field
The invention belongs to the technical field of intelligent power utilization, and particularly relates to a non-invasive deep decomposition method and system for small and miniature load space-time power utilization behaviors.
Background
The existing load decomposition technology can be mainly divided into an invasive type and a non-invasive type. Intrusive Load Monitoring (ILM) is provided with a data acquisition device for each electrical equipment, so that behaviors and energy consumption conditions of different electrical equipment can be accurately known, but the method is high in cost and complex in operation, and is not beneficial to large-scale popularization. Non-Intrusive Load Monitoring (NILM) only installs a data acquisition device at the electrical upstream house entry, and obtains the electricity utilization behavior of the electrical downstream single electrical equipment by Monitoring the aggregated electrical quantity and using a Load decomposition algorithm. Compared with intrusive load decomposition, the non-intrusive load decomposition is relatively difficult, but the non-intrusive method has the advantages of low cost, simplicity in operation and the like, and by means of rapid development of a sensing technology, an information technology and the like, the non-intrusive method also gradually becomes the focus of attention and research of people.
The existing non-invasive load decomposition technology only installs a collecting device with a similar ammeter function at the home-entering position of the electric upstream, and only can identify the time characteristics of the behavior of the electric appliance, but cannot deeply sense the electrical branch and the air characteristics of the electric branch.
Disclosure of Invention
The technical problem to be solved by the invention is to overcome the defects in the prior art and provide a non-invasive deep decomposition method and a non-invasive deep decomposition system for the time-space electricity consumption behavior of the small miniature load so as to identify and obtain the time-scale and space-scale electricity consumption behavior of the small miniature load and deeply sense the time-space electricity consumption behavior of an electric appliance in real time.
In order to achieve the purpose of the invention, the invention adopts a technical scheme that: the non-invasive deep decomposition method for the time-space electric behavior of the small miniature load comprises the following steps:
step 6, electric appliance matching: and (3) substituting the characteristics of the electric appliance to be identified into the probability density functions of the various clusters determined in the steps (1) and (2), determining the type of the electric appliance to be identified, and outputting the type of the electric appliance to be identified and the generation time and space of the power consumption behavior of the electric appliance.
The electrical event in the present invention refers to an on or off event of an electrical appliance, i.e., an electrical appliance disconnection event.
As a further complement to the above technical solution:
in step 1, using active power and reactive power as historical load electricity consumption data characteristics, performing cluster analysis on active power and reactive power variation of historical electrical events to obtain clusters with different numerical values and spatial distribution characteristics, and performing multivariate normal distribution fitting on each cluster generated by clustering to obtain a probability density function:
wherein x is [ P, Q ═ Q]TP, Q are the active power and the reactive power, respectively, and μ and Σ are the mean vector and the covariance matrix of the random variable x, respectively, which are obtained by maximum likelihood estimation as shown in the following equation:
wherein, XiRepresenting the collected historical electricity consumption data samples, N is the total number of the samples,
respectively representing the mean vector and the estimated value of the covariance matrix of the random variable x.
In step 2, calculating according to the corresponding mean vector and covariance matrix estimated for each cluster in step 1
Probability density p of time correspondencerAs per cluster edge probability density, xrRepresenting sample points on the edge of each cluster type;
representing an estimate of the variance of the sample.
In step 3, defining the active power and the reactive power acquired by the main line and each branch of the intelligent circuit breaker as P0,P1,…,PkAnd Q0,Q1,…,QkIn which P is0、Q0Respectively representing active power and reactive power acquired by the trunk line, Pk、QkRespectively representing the active power and the reactive power acquired by the kth branch.
In step 4, the active power of the main circuit collected by the intelligent circuit breaker is subjected to electric appliance event detection, and when the difference value of the active power of two adjacent seconds is more than or equal to a threshold value c1When the event is started, judging that the event of the electric appliance is started; when the difference value of the active power of two adjacent seconds is less than a threshold value c2If the time is longer than the preset time and lasts for three seconds, judging that the electric event is ended, and recording the starting time t and the ending time t of the electric eventstart、tend。
In step 5, calculating the characteristics delta P of the main circuit and each branch circuit event collected by the intelligent circuit breaker from the beginning to the end of the electric appliance eventj＝Pj,end-Pj,start,ΔQj＝Qj,end-Qj,startWhere j is 0,1, …, k, recorded in a matrix, denoted as PA＝[ΔP0,ΔP1,…ΔPk],QA＝[ΔQ0,ΔQ1,…ΔQk]；
In each of the legs 1 to k, the distance Dis ═ Δ P is calculatedj-ΔP0|+|ΔQj-ΔQ0And if j is 0,1, …, k, when the distance Dis takes the minimum value, recording that the branch is m, and determining that the branch is the branch where the electrical event occurs.
In step 6, the characteristics of the electrical appliance to be identified are brought into the probability density functions established in the steps 1 and 2, the probability density p of the electrical appliance is calculated, and when p is reached>prIf the number of the electrical appliances to be identified is r, the time and the space (the branch where the electrical appliance is located) of the type of the electrical appliance to be identified and the electricity consumption behavior thereof are output.
The other technical scheme adopted by the invention is as follows: a system for non-invasive deep decomposition of electrical behavior in time and space under a small miniature load, comprising:
historical load power consumption data acquisition unit: performing cluster analysis on historical load electricity consumption data characteristics of each branch circuit which is normally operated by a user by using an intelligent circuit breaker with a data acquisition function, and performing multivariate normal distribution fitting on each obtained cluster to obtain a probability density function of the cluster;
historical load electricity consumption data analysis unit: calculating the edge probability density of each cluster by using the probability density function obtained by the historical load electricity consumption data acquisition unit;
the real-time load electricity consumption data acquisition unit acquires real-time load electricity consumption data, wherein the real-time load electricity consumption data comprises active power and reactive power per second of a main circuit and each branch circuit acquired by an intelligent circuit breaker;
an electrical event detection unit: judging whether an electrical event starts or ends, if so, turning to the step 5; otherwise, circularly waiting;
an event space-time feature extraction unit: determining a branch where an electrical appliance with an electrical event occurs and calculating the characteristics of the electrical appliance to be identified of the electrical event;
an electric appliance matching unit: and substituting the characteristics of the electric appliance to be identified into various cluster probability density functions determined by the historical load electricity consumption data acquisition unit and the historical load electricity consumption data analysis unit, determining the type of the electric appliance to be identified, and outputting the type of the electric appliance to be identified and the occurrence time and space of the electricity consumption behavior thereof.
Further, in the historical load power consumption data acquisition unit, active power and reactive power are used as historical load power consumption data characteristics, cluster analysis is carried out on the active power and reactive power variation of historical electrical events, clusters with different numerical values and spatial distribution characteristics are obtained, multivariate normal distribution fitting is carried out on each cluster generated by clustering, and a probability density function is obtained:
wherein x is [ P, Q ═ Q]TP, Q are active and reactive power, respectively, and μ and Σ are the mean vector and covariance matrix, respectively, of a random variable x, respectively, byThe maximum likelihood estimate is found as follows:
wherein, XiRepresenting the collected historical electricity consumption data samples, N is the total number of the samples,
respectively representing the mean vector and the estimated value of the covariance matrix of the random variable x.
In the historical load electricity consumption data analysis unit, calculating according to corresponding mean vector and covariance matrix estimated for each cluster in the historical load electricity consumption data acquisition unit
Probability density p of time correspondencerFor each class of cluster edge probability density, xrRepresenting sample points on the edge of each cluster type;
representing an estimate of the variance of the sample.
Further, in the real-time load electricity data acquisition unit, defining the active power and the reactive power acquired by the intelligent breaker trunk circuit and each branch circuit as P0,P1,…,PkAnd Q0,Q1,…,QkIn which P is0、Q0Respectively representing active power and reactive power acquired by the trunk line, Pk、QkRespectively representing active power and reactive power acquired by the kth branch;
in the electric appliance event detection unit, the electric appliance event detection is carried out on the active power of the main circuit collected by the intelligent circuit breaker, and when the difference value of the active power of two adjacent seconds is more than or equal to a threshold value c1When the event is started, judging that the event of the electric appliance is started; when the difference value of the active power of two adjacent seconds is less than a threshold value c2If the time is long and the time lasts for three seconds, the end of the electrical event is judged, and the electrical event is recordedThe occurrence of the start and end times tstart、tend；
In an event space-time characteristic extraction unit, calculating event characteristics delta P of a main circuit and each branch circuit collected by an intelligent breaker from the beginning to the end of an electrical appliance eventj＝Pj,end-Pj,start,ΔQj＝Qj,end-Qj,startWhere j is 0,1, …, k, recorded in a matrix, denoted as PA＝[ΔP0,ΔP1,…ΔPk],QA＝[ΔQ0,ΔQ1,…ΔQk]；
In each of the legs 1 to k, the distance Dis ═ Δ P is calculatedj-ΔP0|+|ΔQj-ΔQ0If j is 0,1, …, k, when the distance Dis takes the minimum value, recording that the branch is m, and determining that m is the branch where the electrical event occurs;
in the electric appliance matching unit, the characteristics of the electric appliance to be identified are brought into the probability density function established by the historical load electricity consumption data acquisition unit and the historical load electricity consumption data analysis unit, the probability density p is calculated, and when p is>prIf the number of the electric appliances to be identified is r, outputting the type of the electric appliances to be identified and the time and space of the electric appliance to be identified.
The invention has the following beneficial technical effects: the invention collects the active power and the reactive power of the small micro load based on the intelligent circuit breaker which has the functions of collecting the electricity data of the main circuit and the multiple circuits, detects the event of the electric appliance, identifies the type of the electric appliance, the action time of the electric appliance and the branch circuit where the electric appliance is located, thereby identifying and obtaining the electricity consumption behavior of the small micro load in time scale and space scale, has the capability of deeply sensing the time-space electricity consumption behavior of a user in real time, and can be used in the field of intelligent electricity consumption of an electric power system.
Drawings
FIG. 1 is a flow chart of a non-invasive deep decomposition method for the time-space electric behavior of a small micro load according to the invention;
FIG. 2 is a clustering effect diagram of the non-invasive deep decomposition method of the small micro load space-time power consumption behavior of the present invention;
FIG. 3 is a block diagram of the non-invasive deep decomposition system for the electric behavior in time and space with small micro load.
Detailed Description
The invention will be further elucidated and described with reference to the drawings and examples. The technical features of the embodiments of the present invention can be combined correspondingly without mutual conflict.
Example 1
A non-invasive deep decomposition method for the time-space electric behavior of a small miniature load is shown in figure 1 and comprises the following steps:
wherein x is [ P, Q ═ Q]Tμ and Σ are the mean vector and covariance matrix of the random variable x, respectively, and can be found by maximum likelihood estimation, as shown in the following equation:
wherein XiRepresenting the collected historical electricity consumption data samples, N is the total number of the samples,
respectively representing the mean vector and the estimated value of the covariance matrix of the random variable x.
Step 6, electric appliance matching: substituting the characteristics of the electrical appliance to be identified into the probability density function established in the steps 1 and 2, calculating the probability density p of the electrical appliance, and when p is>prIf the number of the electrical appliances to be identified is r, the time and the space (the branch where the electrical appliance is located) of the type of the electrical appliance to be identified and the electricity consumption behavior thereof are output.
Application example
The invention provides a method for verifying the effectiveness of a small micro space-time electricity consumption behavior non-invasive deep decomposition method. Based on a mode of combining a data terminal and a cloud platform, an intelligent circuit breaker produced by a certain company is used for data acquisition and uploading to the cloud platform, and a calculation and processing function is implemented on the cloud platform, so that the test and verification of the embodiment are completed.
A resident user family containing small and miniature loads is provided with an intelligent circuit breaker with a data acquisition function, a main line and five branches. Five branches of the intelligent circuit breaker are respectively connected with five small miniature loads of a refrigerator, a water heater, a microwave oven, an electric cooker and a hot water kettle. The historical electricity consumption data is sorted, and a clustering effect graph obtained by clustering analysis is shown in the attached figure 2.
The starting and stopping tests of the five types of household electrical appliances are carried out to obtain the actual event number and the identification event number, and the identification accuracy and the branch result of the electrical appliance are shown in table 1.
TABLE 1
| Refrigerator with a door | Hot water kettle | Water heater | Microwave oven with a heat exchanger | Electric rice cooker | |
| Identifying number of |
50 | 62 | 58 | 70 | 85 |
| Actual number of events | 53 | 65 | 62 | 77 | 90 |
| Rate of accuracy | 94.3％ | 95.4％ | 93.5％ | 90.9％ | 94.4 |
| Branch circuit | |||||
| 1 | 5 | 2 | 3 | 4 |
According to the result of the graph, the algorithm has good identification effect, and non-invasive deep decomposition of the space-time power consumption behavior of the small and micro load can be realized.
Example 2
A system for non-invasive deep decomposition of spatiotemporal electrical behavior of small micro-load as shown in fig. 3, comprising:
historical load power consumption data acquisition unit: performing cluster analysis on historical load electricity consumption data characteristics of each branch circuit which is normally operated by a user by using an intelligent circuit breaker with a data acquisition function, and performing multivariate normal distribution fitting on each obtained cluster to obtain a probability density function of the cluster;
historical load electricity consumption data analysis unit: calculating the edge probability density of each cluster by using the probability density function obtained by the historical load electricity consumption data acquisition unit;
the real-time load electricity consumption data acquisition unit acquires real-time load electricity consumption data, wherein the real-time load electricity consumption data comprises active power and reactive power per second of a main circuit and each branch circuit acquired by an intelligent circuit breaker;
an electrical event detection unit: judging whether an electrical event starts or ends, if so, turning to the step 5; otherwise, circularly waiting;
an event space-time feature extraction unit: determining a branch where an electrical appliance with an electrical event occurs and calculating the characteristics of the electrical appliance to be identified of the electrical event;
an electric appliance matching unit: and substituting the characteristics of the electric appliance to be identified into various cluster probability density functions determined by the historical load electricity consumption data acquisition unit and the historical load electricity consumption data analysis unit, determining the type of the electric appliance to be identified, and outputting the type of the electric appliance to be identified and the occurrence time and space of the electricity consumption behavior thereof.
In a historical load electricity consumption data acquisition unit, using active power and reactive power as historical load electricity consumption data characteristics, carrying out cluster analysis on the active power and reactive power variation of historical electrical events of the historical electrical events to obtain clusters with different numerical values and spatial distribution characteristics, and carrying out multivariate normal distribution fitting on each cluster generated by clustering to obtain a probability density function:
wherein x is [ P, Q ═ Q]TP, Q are active and reactive power, respectively, and μ and Σ are random, respectivelyThe mean vector and covariance matrix of variable x are obtained by maximum likelihood estimation as shown in the following formula:
wherein, XiRepresenting the collected historical electricity consumption data samples, N is the total number of the samples,
respectively representing the mean vector and the estimated value of the covariance matrix of the random variable x.
In the historical load electricity consumption data analysis unit, calculating according to corresponding mean vector and covariance matrix estimated for each cluster in the historical load electricity consumption data acquisition unit
Probability density p of time correspondencerAs per cluster edge probability density, xrRepresenting sample points on the edge of each cluster type;
representing an estimate of the variance of the sample.
In a real-time load electricity data acquisition unit, defining active power and reactive power acquired by a main circuit and each branch of an intelligent circuit breaker as P0,P1,…,PkAnd Q0,Q1,…,QkIn which P is0、Q0Respectively representing active power and reactive power acquired by the trunk line, Pk、QkRespectively representing active power and reactive power acquired by the kth branch;
in the electric appliance event detection unit, the electric appliance event detection is carried out on the active power of the main circuit collected by the intelligent circuit breaker, and when the difference value of the active power of two adjacent seconds is more than or equal to a threshold value c1When the event is started, judging that the event of the electric appliance is started; when the difference value of the active power of two adjacent seconds is less than a threshold value c2When the time is long and lasts for three seconds, the electricity is judgedEnding the event, recording the time t of the beginning and ending of the eventstart、tend；
In an event space-time characteristic extraction unit, calculating event characteristics delta P of a main circuit and each branch circuit collected by an intelligent breaker from the beginning to the end of an electrical appliance eventj＝Pj,end-Pj,start,ΔQj＝Qj,end-Qj,startWhere j is 0,1, …, k, recorded in a matrix, denoted as PA＝[ΔP0,ΔP1,…ΔPk],QA＝[ΔQ0,ΔQ1,…ΔQk]；
In each of the legs 1 to k, the distance Dis ═ Δ P is calculatedj-ΔP0|+|ΔQj-ΔQ0If j is 0,1, …, k, when the distance Dis takes the minimum value, recording that the branch is m, and determining that m is the branch where the electrical event occurs;
in the electric appliance matching unit, the characteristics of the electric appliance to be identified are brought into the probability density function established by the historical load electricity consumption data acquisition unit and the historical load electricity consumption data analysis unit, the probability density p is calculated, and when p is>prIf the number of the electric appliances to be identified is r, outputting the type of the electric appliances to be identified and the time and space of the electric appliance to be identified.
Any alternatives, modifications and variations are within the spirit and scope of the invention as defined by the appended claims.
Claims (10)
1. The non-invasive deep decomposition method for the time-space electric behavior of the small and miniature load is characterized by comprising the following steps:
step 1, collecting historical load electricity consumption data: performing cluster analysis on historical load electricity consumption data characteristics of each branch circuit which is normally operated by a user by using an intelligent circuit breaker with a data acquisition function, and performing multivariate normal distribution fitting on each obtained cluster to obtain a probability density function of the cluster;
step 2, analyzing historical load electricity consumption data: calculating the edge probability density of each cluster by using the probability density function obtained in the step 1;
step 3, collecting real-time load electricity utilization data, wherein the real-time load electricity utilization data comprise active power and reactive power per second of a main circuit and each branch circuit collected by an intelligent circuit breaker;
step 4, detecting an electric appliance event: judging whether an electrical event starts or ends, if so, turning to the step 5; otherwise, circularly waiting;
step 5, event spatio-temporal feature extraction: determining a branch where an electrical appliance with an electrical event occurs and calculating the characteristics of the electrical appliance to be identified of the electrical event;
step 6, electric appliance matching: and (3) substituting the characteristics of the electric appliance to be identified into the probability density functions of the various clusters determined in the steps (1) and (2), determining the type of the electric appliance to be identified, and outputting the type of the electric appliance to be identified and the generation time and space of the power consumption behavior of the electric appliance.
2. The small micro-load space-time power consumption behavior non-invasive deep decomposition method according to claim 1, characterized in that in step 1, active power and reactive power are used as historical load power consumption data characteristics, cluster analysis is performed on active power and reactive power variation of historical electrical events of the small micro-load space-time power consumption data characteristics, clusters with different numerical values and spatial distribution characteristics are obtained, multivariate normal distribution fitting is performed on each cluster generated by clustering, and a probability density function is obtained:
wherein x is [ P, Q ═ Q]TP, Q are the active power and the reactive power, respectively, and μ and Σ are the mean vector and the covariance matrix of the random variable x, respectively, obtained by maximum likelihood estimation as shown in the following equation:
wherein, XiRepresenting the collected historical electricity consumption data samples, N is the total number of the samples,
respectively representing the mean vector and the estimated value of the covariance matrix of the random variable x.
3. The method according to claim 2, wherein in step 2, the calculation is performed according to the corresponding mean vector and covariance matrix estimated for each cluster in step 1
Probability density p of time correspondencerAs per cluster edge probability density, xrRepresenting sample points on the edge of each cluster type;
representing an estimate of the variance of the sample.
4. The non-invasive deep decomposition method for small micro-load space-time power consumption behaviors according to claim 3, wherein in the step 3, active power and reactive power collected by a main circuit and each branch of the intelligent circuit breaker are defined as P0,P1,…,PkAnd Q0,Q1,…,QkIn which P is0、Q0Respectively representing active power and reactive power acquired by the trunk line, Pk、QkRespectively representing the active power and the reactive power acquired by the kth branch.
5. The non-invasive deep decomposition method for small micro load space-time electric behavior according to claim 1, wherein in step 4, the electric event detection is performed on the main circuit active power collected by the intelligent circuit breaker, and when the difference between two adjacent seconds of the active power is greater than or equal to a threshold value c1When the event is started, judging that the event of the electric appliance is started; when the difference value of the active power of two adjacent seconds is less than a threshold value c2If the time is long and the time lasts for three seconds, the end of the electric appliance event is judged, and the electric appliance event is recordedPiece occurrence start and end time tstart、tend。
6. The method for the non-invasive deep decomposition of the spatio-temporal electric behavior of the small miniature load according to claim 4, wherein in step 5, the characteristics Δ P of the main circuit and each branch circuit event collected by the intelligent circuit breaker from the beginning to the end of the electric event are calculatedj＝Pj,end-Pj,start,ΔQj＝Qj,end-Qj,startWhere j is 0,1, …, k, recorded in a matrix, denoted as PA＝[ΔP0,ΔP1,…ΔPk],QA＝[ΔQ0,ΔQ1,…ΔQk]；
In each of the legs 1 to k, the distance Dis ═ Δ P is calculatedj-ΔP0|+|ΔQj-ΔQ0And if j is 0,1, …, k, when the distance Dis takes the minimum value, recording that the branch is m, and determining that the branch is the branch where the electrical event occurs.
7. The method for the non-invasive deep decomposition of the electric behavior in time and space of the small micro load according to claim 6, wherein in step 6, the characteristics of the electric appliance to be identified are brought into the probability density functions established in the steps 1 and 2, the probability density p is calculated, and when p is reached, the probability density p is calculated>prIf the number of the electric appliances to be identified is r, outputting the type of the electric appliances to be identified and the time and space of the electric appliance to be identified.
8. Little miniature load space-time power consumption action non-invasive degree of depth decomposition system, its characterized in that includes:
historical load power consumption data acquisition unit: performing cluster analysis on historical load electricity consumption data characteristics of each branch circuit which is normally operated by a user by using an intelligent circuit breaker with a data acquisition function, and performing multivariate normal distribution fitting on each obtained cluster to obtain a probability density function of the cluster;
historical load electricity consumption data analysis unit: calculating the edge probability density of each cluster by using the probability density function obtained by the historical load electricity consumption data acquisition unit;
the real-time load electricity consumption data acquisition unit acquires real-time load electricity consumption data, wherein the real-time load electricity consumption data comprises active power and reactive power per second of a main circuit and each branch circuit acquired by an intelligent circuit breaker;
an electrical event detection unit: judging whether an electrical event starts or ends, if so, turning to the step 5; otherwise, circularly waiting;
an event space-time feature extraction unit: determining a branch where an electrical appliance with an electrical event occurs and calculating the characteristics of the electrical appliance to be identified of the electrical event;
an electric appliance matching unit: and substituting the characteristics of the electric appliance to be identified into various cluster probability density functions determined by the historical load electricity consumption data acquisition unit and the historical load electricity consumption data analysis unit, determining the type of the electric appliance to be identified, and outputting the type of the electric appliance to be identified and the occurrence time and space of the electricity consumption behavior thereof.
9. The system according to claim 8, wherein in the historical load electricity data collection unit, active power and reactive power are used as historical load electricity data characteristics, cluster analysis is performed on the active power and reactive power variation of historical electrical events to obtain clusters with different values and spatial distribution characteristics, and multivariate normal distribution fitting is performed on each cluster generated by clustering to obtain a probability density function:
wherein x is [ P, Q ═ Q]TP, Q are the active power and the reactive power, respectively, and μ and Σ are the mean vector and the covariance matrix of the random variable x, respectively, obtained by maximum likelihood estimation as shown in the following equation:
wherein the content of the first and second substances,Xirepresenting the collected historical electricity consumption data samples, N is the total number of the samples,
respectively representing the mean vector of the random variable x and the estimated value of the covariance matrix;
in the historical load electricity consumption data analysis unit, calculating according to corresponding mean vector and covariance matrix estimated for each cluster in the historical load electricity consumption data acquisition unit
Probability density p of time correspondencerAs per cluster edge probability density, xrRepresenting sample points on the edge of each cluster type;
representing an estimate of the variance of the sample.
10. The system according to claim 9, wherein the real-time load power consumption data collection unit defines the active power and the reactive power collected by the intelligent breaker trunk and each branch as P0,P1,…,PkAnd Q0,Q1,…,QkIn which P is0、Q0Respectively representing active power and reactive power acquired by the trunk line, Pk、QkRespectively representing active power and reactive power acquired by the kth branch;
in the electric appliance event detection unit, the electric appliance event detection is carried out on the active power of the main circuit collected by the intelligent circuit breaker, and when the difference value of the active power of two adjacent seconds is more than or equal to a threshold value c1When the event is started, judging that the event of the electric appliance is started; when the difference value of the active power of two adjacent seconds is less than a threshold value c2If the time is longer than the preset time and lasts for three seconds, judging that the electric event is ended, and recording the starting time t and the ending time t of the electric eventstart、tend；
In an event space-time characteristic extraction unit, calculating event characteristics delta P of a main circuit and each branch circuit collected by an intelligent breaker from the beginning to the end of an electrical appliance eventj＝Pj,end-Pj,start,ΔQj＝Qj,end-Qj,startWhere j is 0,1, …, k, recorded in a matrix, denoted as PA＝[ΔP0,ΔP1,…ΔPk],QA＝[ΔQ0,ΔQ1,…ΔQk]；
In each of the legs 1 to k, the distance Dis ═ Δ P is calculatedj-ΔP0|+|ΔQj-ΔQ0If j is 0,1, …, k, when the distance Dis takes the minimum value, recording that the branch is m, and determining that m is the branch where the electrical event occurs;
in the electric appliance matching unit, the characteristics of the electric appliance to be identified are brought into the probability density function established by the historical load electricity consumption data acquisition unit and the historical load electricity consumption data analysis unit, the probability density p is calculated, and when p is>prIf the number of the electric appliances to be identified is r, outputting the type of the electric appliances to be identified and the time and space of the electric appliance to be identified.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110405550.1A CN113269227A (en) | 2021-04-15 | 2021-04-15 | Non-invasive deep decomposition method and system for small and miniature load time-space electricity consumption behavior |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110405550.1A CN113269227A (en) | 2021-04-15 | 2021-04-15 | Non-invasive deep decomposition method and system for small and miniature load time-space electricity consumption behavior |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113269227A true CN113269227A (en) | 2021-08-17 |
Family
ID=77228880
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110405550.1A Pending CN113269227A (en) | 2021-04-15 | 2021-04-15 | Non-invasive deep decomposition method and system for small and miniature load time-space electricity consumption behavior |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113269227A (en) |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150309092A1 (en) * | 2012-11-16 | 2015-10-29 | Tianjin University | Current Pattern Matching Method for Non-Intrusive Power Load Monitoring and Disaggregation |
| US20160156225A1 (en) * | 2014-12-01 | 2016-06-02 | Eaton Corporation | Load power device and system for real-time execution of hierarchical load identification algorithms |
| CN108694470A (en) * | 2018-06-12 | 2018-10-23 | 天津大学 | A kind of data predication method and device based on artificial intelligence |
| CN110472777A (en) * | 2019-07-29 | 2019-11-19 | 上海电力大学 | A kind of micro-capacitance sensor short-term load forecasting method based on empirical mode decomposition |
| CN110689190A (en) * | 2019-09-24 | 2020-01-14 | 南方电网科学研究院有限责任公司 | Power grid load prediction method and device and related equipment |
| CN110907762A (en) * | 2019-12-10 | 2020-03-24 | 深圳供电局有限公司 | Non-invasive load matching identification method |
| CN111553444A (en) * | 2020-05-15 | 2020-08-18 | 东南大学 | Load identification method based on non-invasive load terminal data |
| CN111682643A (en) * | 2020-06-04 | 2020-09-18 | 国网浙江杭州市萧山区供电有限公司 | Intelligent circuit breaker capable of realizing power load characteristic identification |
| CN111934318A (en) * | 2020-08-13 | 2020-11-13 | 彭浩明 | Non-invasive power load decomposition method, device, equipment and storage medium |
| CN112311088A (en) * | 2019-07-29 | 2021-02-02 | 南寅 | Electric power and home information monitoring system, method, equipment and medium |
| CN112505511A (en) * | 2020-12-14 | 2021-03-16 | 天津求实智源科技有限公司 | Non-invasive low-voltage fault arc detection and positioning method and system |
-
2021
- 2021-04-15 CN CN202110405550.1A patent/CN113269227A/en active Pending
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150309092A1 (en) * | 2012-11-16 | 2015-10-29 | Tianjin University | Current Pattern Matching Method for Non-Intrusive Power Load Monitoring and Disaggregation |
| US20160156225A1 (en) * | 2014-12-01 | 2016-06-02 | Eaton Corporation | Load power device and system for real-time execution of hierarchical load identification algorithms |
| CN108694470A (en) * | 2018-06-12 | 2018-10-23 | 天津大学 | A kind of data predication method and device based on artificial intelligence |
| CN110472777A (en) * | 2019-07-29 | 2019-11-19 | 上海电力大学 | A kind of micro-capacitance sensor short-term load forecasting method based on empirical mode decomposition |
| CN112311088A (en) * | 2019-07-29 | 2021-02-02 | 南寅 | Electric power and home information monitoring system, method, equipment and medium |
| CN110689190A (en) * | 2019-09-24 | 2020-01-14 | 南方电网科学研究院有限责任公司 | Power grid load prediction method and device and related equipment |
| CN110907762A (en) * | 2019-12-10 | 2020-03-24 | 深圳供电局有限公司 | Non-invasive load matching identification method |
| CN111553444A (en) * | 2020-05-15 | 2020-08-18 | 东南大学 | Load identification method based on non-invasive load terminal data |
| CN111682643A (en) * | 2020-06-04 | 2020-09-18 | 国网浙江杭州市萧山区供电有限公司 | Intelligent circuit breaker capable of realizing power load characteristic identification |
| CN111934318A (en) * | 2020-08-13 | 2020-11-13 | 彭浩明 | Non-invasive power load decomposition method, device, equipment and storage medium |
| CN112505511A (en) * | 2020-12-14 | 2021-03-16 | 天津求实智源科技有限公司 | Non-invasive low-voltage fault arc detection and positioning method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Altrabalsi et al. | Low-complexity energy disaggregation using appliance load modelling | |
| Devlin et al. | Non-intrusive load monitoring and classification of activities of daily living using residential smart meter data | |
| CN106936129B (en) | Power load identification method and system based on multi-feature fusion | |
| CN111564842B (en) | Method for statistical estimation of medium-and-long-term energy consumption in non-invasive electric load monitoring | |
| Berges et al. | Learning systems for electric consumption of buildings | |
| CN110504679B (en) | KM matching algorithm-based non-invasive load identification method | |
| Wang et al. | Residential appliances identification and monitoring by a nonintrusive method | |
| CN111244954A (en) | Non-invasive load identification method and device | |
| CN108021736A (en) | A kind of load switching action monitoring method based on sliding window Remanent Model | |
| Liang et al. | HVAC load disaggregation using low-resolution smart meter data | |
| Devlin et al. | Non-intrusive load monitoring using electricity smart meter data: A deep learning approach | |
| CN112505511B (en) | Non-invasive low-voltage fault arc detection and positioning method and system | |
| CN106646026A (en) | Non-intrusive household appliance load identification method | |
| CN113361831B (en) | Non-invasive load identification electric quantity decomposition method and system based on priority distribution | |
| Girmay et al. | Simple event detection and disaggregation approach for residential energy estimation | |
| CN116008714B (en) | Anti-electricity-stealing analysis method based on intelligent measurement terminal | |
| CN112909923A (en) | Non-invasive household load behavior recognition device based on DTW algorithm | |
| CN115456034A (en) | Automatic identification and monitoring method and system for electric bicycle charging | |
| Welikala et al. | Robust non-intrusive load monitoring (nilm) with unknown loads | |
| Luan et al. | Non-intrusive power waveform modeling and identification of air conditioning load | |
| Chen et al. | Cloud-based non-intrusive leakage current detection for residential appliances | |
| CN113269227A (en) | Non-invasive deep decomposition method and system for small and miniature load time-space electricity consumption behavior | |
| Devlin et al. | Load identification and classification of activities of daily living using residential smart meter data | |
| CN205809165U (en) | A kind of load identification system for domestic intelligent electricity meter based on similarity algorithm | |
| Dinesh et al. | Non-intrusive load monitoring based on low frequency active power measurements |
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: 20210817 |