CN115564324A - Electric power fingerprint identification method, device and equipment based on event detection - Google Patents
Electric power fingerprint identification method, device and equipment based on event detection Download PDFInfo
- Publication number
- CN115564324A CN115564324A CN202211175286.8A CN202211175286A CN115564324A CN 115564324 A CN115564324 A CN 115564324A CN 202211175286 A CN202211175286 A CN 202211175286A CN 115564324 A CN115564324 A CN 115564324A
- Authority
- CN
- China
- Prior art keywords
- data
- load
- edge distance
- current
- event
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 51
- 238000000034 method Methods 0.000 title claims abstract description 50
- 238000005070 sampling Methods 0.000 claims abstract description 37
- 238000012549 training Methods 0.000 claims description 58
- 230000002776 aggregation Effects 0.000 claims description 23
- 238000004220 aggregation Methods 0.000 claims description 23
- 230000000630 rising effect Effects 0.000 claims description 16
- 238000007781 pre-processing Methods 0.000 claims description 13
- 230000002159 abnormal effect Effects 0.000 claims description 9
- 238000007637 random forest analysis Methods 0.000 claims description 8
- 238000002372 labelling Methods 0.000 claims description 5
- 238000004590 computer program Methods 0.000 claims description 4
- 238000000605 extraction Methods 0.000 description 22
- 230000008569 process Effects 0.000 description 8
- 238000010276 construction Methods 0.000 description 7
- 230000002708 enhancing effect Effects 0.000 description 6
- 238000004364 calculation method Methods 0.000 description 4
- 238000003066 decision tree Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000000926 separation method Methods 0.000 description 3
- 230000001052 transient effect Effects 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000013480 data collection Methods 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000006116 polymerization reaction Methods 0.000 description 2
- YTAHJIFKAKIKAV-XNMGPUDCSA-N [(1R)-3-morpholin-4-yl-1-phenylpropyl] N-[(3S)-2-oxo-5-phenyl-1,3-dihydro-1,4-benzodiazepin-3-yl]carbamate Chemical compound O=C1[C@H](N=C(C2=C(N1)C=CC=C2)C1=CC=CC=C1)NC(O[C@H](CCN1CCOCC1)C1=CC=CC=C1)=O YTAHJIFKAKIKAV-XNMGPUDCSA-N 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 238000013145 classification model Methods 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- Development Economics (AREA)
- Health & Medical Sciences (AREA)
- Educational Administration (AREA)
- Marketing (AREA)
- Entrepreneurship & Innovation (AREA)
- Theoretical Computer Science (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Game Theory and Decision Science (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of load identification, and discloses a power fingerprint identification method, device and equipment based on event detection. Sampling electrical quantity data by adopting a sliding window with multistage edge distance windows at two ends, further respectively calculating the average value of the sampled electrical quantities of a main edge distance window at the left end of the sliding window, a secondary edge distance window at the left end, a main edge distance window at the right end and a secondary edge distance window at the right end, and if the absolute difference values of the average values of the electrical quantities corresponding to the main edge distance window and the secondary edge distance windows at the two ends of the sliding window are all larger than a preset electrical quantity threshold value and the time interval of the current sampling time to the previous load switch event is larger than a preset time threshold value, judging that a load switch event occurs and extracting corresponding load waveform data; and extracting power fingerprint characteristics according to the extracted load waveform data corresponding to each load switch event, inputting the power fingerprint characteristics to a trained power fingerprint identification model, and outputting a load type identification result. The invention can effectively improve the identification accuracy of the power fingerprint identification.
Description
Technical Field
The invention relates to the technical field of load identification, in particular to a power fingerprint identification method, device and equipment based on event detection.
Background
By monitoring the electrical data of the electrical equipment, feature points capable of representing the characteristics of the equipment are mined by using an artificial intelligence technology and a big data technology, and the feature points of all dimensions are aggregated to form the 'power fingerprint' of the equipment. The electric power fingerprint technology further expands the connotation of the load characteristics on the basis of the research results of the existing load characteristics and identification methods, and can be used for identifying the type, characteristics, parameters, user behavior habits, energy efficiency, health level and identity of equipment.
Electric fingerprint identification can be divided into two categories: non-event based identification and event based identification. Directly assigning aggregated characteristics of the bus to each electrical device non-based on event identification; based on event identification, attention is paid to sudden change of electrical characteristics of a bus caused by state switching of load equipment (also called load events), and the electrical equipment is identified by detecting and extracting a sudden change quantity. Compared with non-event-based identification, the real-time performance and the overall identification accuracy rate of the event-based identification are better.
However, when the power fingerprint identification is performed based on the event detection, the long transient state and the generated surge current when some loads are connected, and the abnormal value fluctuation condition occurring when the monitoring equipment is used for measurement, can cause the event false detection, and the overall identification accuracy is reduced. In addition, when the robustness of the power fingerprint identification model is low due to fewer current load samples, the identification accuracy of the load under the disturbance of the combined load is also reduced.
Disclosure of Invention
The invention provides a power fingerprint identification method, a device and equipment based on event detection, and solves the technical problem that the accuracy identification rate of the existing power fingerprint identification method based on event detection is still to be improved.
The invention provides a power fingerprint identification method based on event detection, which comprises the following steps:
sampling electrical quantity data by adopting a sliding window with multistage edge distance windows arranged at two ends; the length of the sliding window is greater than the sum of the lengths of the multistage edge distance windows at the two ends of the sliding window, and the multistage edge distance window at each end consists of a main edge distance window and a secondary edge distance window;
after sampling each time, respectively calculating the average values of the sampled electrical quantities of the main margin window at the left end of the sliding window, the secondary margin window at the left end, the main margin window at the right end and the secondary margin window at the right end, if the absolute values of the difference values of the electrical quantities corresponding to the main margin windows at the two ends of the sliding window and the secondary margin windows are all larger than a preset electrical quantity threshold value, and the time interval between the current sampling time and the previous load switch event is larger than the preset time threshold value, judging that the load switch event occurs and extracting corresponding load waveform data;
and extracting corresponding electric power fingerprint characteristics according to the extracted load waveform data corresponding to each load switch event, inputting the electric power fingerprint characteristics into a trained electric power fingerprint identification model, and outputting a load type identification result.
According to one possible implementation of the first aspect of the invention, the electrical quantity data is an active power sequence composed of active power of each period, and the active power is calculated from voltage and current measured in the corresponding period.
According to an implementation manner of the first aspect of the present invention, the corresponding load waveform data includes voltage and current data before and after occurrence of a corresponding load switch event, and the extracting corresponding power fingerprint features according to the extracted load waveform data corresponding to each load switch event includes:
segmenting voltage and current data before and after each load switch event by taking a zero crossing point of a voltage rising stage as a boundary, so that each obtained data segment is data of one cycle;
if the load switch event is the load access, subtracting the current data before the corresponding event from the current data section after the corresponding event occurs to obtain separated current data; if the load switch event is load cutting, subtracting the current data after the corresponding event from the current data section before the corresponding event to obtain separated current data;
merging the separated current data with the voltage data section after the corresponding event to obtain separated load data;
and extracting the electric power fingerprint characteristics of each separated load data.
According to a possible implementation manner of the first aspect of the present invention, the extracting the power fingerprint feature of each separated load data includes:
extracting power characteristics, harmonic characteristics and waveform characteristics of the separated load data; the power characteristics include active power, reactive power, apparent power and power factor; the harmonic characteristics comprise effective values and phases of fundamental wave, 2, 3, 5 and 7 times of harmonic current and impedance, and further comprise 4-6-8-10 times of current harmonic aggregation effective values, 9-12-15 times of current harmonic aggregation effective values and 11-13-17-19-23-25 times of current harmonic aggregation effective values; the waveform characteristics include a total harmonic distortion rate, a form factor, and a crest factor.
According to a manner that can be realized by the first aspect of the present invention, the extracting the power feature, the harmonic feature and the waveform feature of the separated load data includes:
calculating the harmonic aggregation effective value of each current according to the following formula:
in the formula I Σ Represents the effective value of current harmonic aggregation, I h Representing the effective value of H harmonic current, H being the set of harmonic orders aggregated.
According to an enabling aspect of the first aspect of the invention, the method further comprises:
collecting and labeling steady state load waveform data of each type of single load, extracting power fingerprint characteristics of corresponding single loads according to the labeled steady state load waveform data, and constructing an initial training set;
performing data enhancement on the initial training set to obtain a training set after data enhancement;
and training the electric power fingerprint identification model by using the training set after the data enhancement to obtain the trained electric power fingerprint identification model.
According to a manner that can be realized by the first aspect of the present invention, the extracting the power fingerprint feature of the corresponding individual load according to the labeled steady-state load waveform data includes:
preprocessing the labeled steady-state load waveform data; the preprocessing comprises filling data missing values, removing abnormal data, removing data which are not matched with corresponding labels and/or removing data which are not matched with corresponding switch states;
and extracting the electric power fingerprint characteristics of the corresponding single load from the preprocessed steady-state load waveform data.
According to an implementable manner of the first aspect of the present invention, the data enhancement of the initial training set includes:
selecting a group of data from the preprocessed steady-state load waveform data, segmenting the data according to a zero crossing point of a voltage rising stage to enable each obtained data segment to be data of a cycle wave, and selecting two unequal current waveforms from each obtained data segment to subtract to obtain corresponding disturbance waveform data;
repeating the previous step to obtain a plurality of disturbance waveform data, and constructing a disturbance waveform data set according to the plurality of disturbance waveform data;
selecting any group of data from the preprocessed steady-state load waveform data, segmenting the data according to a zero crossing point of a voltage rising stage, determining a target current data segment from each obtained data segment, randomly selecting a plurality of disturbance waveform data from the disturbance waveform data set to be superposed on the target current data segment, and obtaining reconstructed load data without changing voltage data and labels corresponding to the target current data segment;
repeating the previous step to obtain a plurality of reconstruction load data, and constructing a reconstruction load data set according to the plurality of reconstruction load data;
and extracting the electric power fingerprint characteristics of the corresponding individual load from each reconstructed load data in the reconstructed load data set, and adding the electric power fingerprint characteristics into the initial training set.
According to an implementable manner of the first aspect of the present invention, training the power fingerprint recognition model using the data-enhanced training set comprises:
and constructing a random forest model as the electric power fingerprint identification model.
The invention provides a power fingerprint identification device based on event detection, which comprises:
the sampling module is used for sampling the electrical quantity data by adopting a sliding window with multistage edge distance windows arranged at two ends; the length of the sliding window is greater than the sum of the lengths of the multistage edge distance windows at the two ends of the sliding window, and the multistage edge distance window at each end consists of a main edge distance window and a secondary edge distance window;
the event detection module is used for respectively calculating the average values of the sampled electrical quantities of the main edge distance window at the left end of the sliding window, the secondary edge distance window at the left end, the main edge distance window at the right end and the secondary edge distance window at the right end after sampling each time, and if the absolute difference values of the electric quantity average values corresponding to the main edge distance windows and the secondary edge distance windows at the two ends of the sliding window are all larger than a preset electrical quantity threshold value and the time interval of the current sampling time to the previous load switch event is larger than the preset time threshold value, judging that the load switch event occurs and extracting corresponding load waveform data;
and the electric power fingerprint identification module is used for extracting corresponding electric power fingerprint characteristics according to the extracted load waveform data corresponding to each load switch event, inputting the electric power fingerprint characteristics into a trained electric power fingerprint identification model and outputting a load type identification result.
According to a manner that can be realized by the second aspect of the present invention, the electrical quantity data is an active power sequence composed of active power of each period, and the active power is calculated by voltage and current measured by the corresponding period.
According to a second aspect of the present invention, in an implementation manner, the corresponding load waveform data includes voltage and current data before and after a corresponding load switch event occurs, and the power fingerprint identification module includes:
the waveform alignment unit is used for segmenting voltage and current data before and after each load switch event by taking a zero crossing point of a voltage rising stage as a boundary so as to enable each obtained data segment to be data of one cycle;
the first increment extraction unit is used for subtracting the current data before the corresponding event from the current data section after the corresponding event occurs to obtain separated current data if the load switch event is the loaded access; if the load switch event is load cutting, subtracting the current data after the corresponding event from the current data section before the corresponding event to obtain separated current data;
the data merging unit is used for merging the separated current data with the voltage data section after the corresponding event to obtain separated load data;
and the first characteristic extraction unit is used for extracting the electric power fingerprint characteristics of each separated load data.
According to an implementable manner of the second aspect of the present invention, the first feature extraction unit includes:
the characteristic extraction subunit is used for extracting the power characteristic, the harmonic characteristic and the waveform characteristic of the separated load data; the power characteristics include active power, reactive power, apparent power and power factor; the harmonic characteristics comprise effective values and phases of fundamental wave, 2, 3, 5 and 7 times of harmonic current and impedance, and further comprise 4-6-8-10 times of current harmonic aggregation effective values, 9-12-15 times of current harmonic aggregation effective values and 11-13-17-19-23-25 times of current harmonic aggregation effective values; the waveform characteristics include a total harmonic distortion rate, a form factor, and a crest factor.
According to an implementable manner of the second aspect of the present invention, the feature extraction subunit is specifically configured to:
calculating the polymerization effective value of each current harmonic according to the following formula:
in the formula I Σ Represents the effective value of current harmonic aggregation, I h Representing the effective value of the H harmonic current, H being the set of harmonic orders aggregated.
According to an implementable manner of the second aspect of the invention, the apparatus further comprises:
the initial training set building module is used for collecting the steady-state load waveform data of each type of single load, labeling the steady-state load waveform data, extracting the power fingerprint characteristics of the corresponding single load according to the labeled steady-state load waveform data, and building an initial training set;
the training set enhancing module is used for enhancing the data of the initial training set to obtain a training set after data enhancement;
and the model training module is used for training the electric power fingerprint identification model by using the training set after the data enhancement to obtain the trained electric power fingerprint identification model.
According to an implementable manner of the second aspect of the present invention, the initial training set constructing module comprises:
the data preprocessing unit is used for preprocessing the labeled steady-state load waveform data; the preprocessing comprises filling data missing values, removing abnormal data, removing data which are not matched with corresponding labels and/or removing data which are not matched with corresponding switch states;
and the second characteristic extraction unit is used for extracting the electric power fingerprint characteristics of the corresponding single load from the preprocessed steady-state load waveform data.
According to one possible implementation of the second aspect of the invention, the training set enhancing module comprises:
the second increment extraction unit is used for selecting a group of data from the preprocessed steady-state load waveform data, segmenting the data according to the zero crossing point of the voltage rising stage so as to enable each obtained data segment to be data of a cycle wave, and selecting two unequal current waveforms from each obtained data segment for subtraction to obtain corresponding disturbance waveform data;
the disturbance waveform data set construction unit is used for repeating the second increment extraction unit to obtain a plurality of disturbance waveform data and constructing a disturbance waveform data set according to the plurality of disturbance waveform data;
a third increment extraction unit, configured to select any one group of data from the preprocessed steady-state load waveform data, segment the data according to a zero crossing point of a voltage rising stage, determine a target current data segment from the obtained data segments, randomly select a plurality of quantities of disturbance waveform data from the disturbance waveform data set, superimpose the disturbance waveform data onto the target current data segment, and obtain reconstructed load data without changing voltage data and a label corresponding to the target current data segment;
the reconstructed load data set construction unit is used for repeating the third increment extraction unit to obtain a plurality of reconstructed load data, and constructing a reconstructed load data set according to the plurality of reconstructed load data;
and the third feature extraction unit is used for extracting the power fingerprint features of the corresponding individual loads from each reconstructed load data in the reconstructed load data set and adding the power fingerprint features into the initial training set.
According to an enabling manner of the second aspect of the present invention, the model training module comprises:
and the model construction unit is used for constructing a random forest model as the electric power fingerprint identification model.
The invention provides a power fingerprint identification device based on event detection, which comprises:
a memory to store instructions; wherein the instructions are for implementing an event detection-based power fingerprinting method as described in any of the above implementable manners;
a processor to execute the instructions in the memory.
A fourth aspect of the present invention is a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements an event detection-based power fingerprinting method as described in any of the above-implementable manners.
According to the technical scheme, the invention has the following advantages:
the invention adopts a sliding window with multi-stage edge distance windows at two ends to sample the electrical quantity data; the length of the sliding window is greater than the sum of the lengths of the multistage edge distance windows at two ends of the sliding window, and the multistage edge distance window at each end consists of a main edge distance window and a secondary edge distance window; after sampling each time, respectively calculating the average values of the sampled electrical quantities of the main margin window at the left end of the sliding window, the secondary margin window at the left end, the main margin window at the right end and the secondary margin window at the right end, if the absolute values of the difference values of the electrical quantities corresponding to the main margin windows at the two ends of the sliding window and the secondary margin windows are all larger than a preset electrical quantity threshold value, and the time interval between the current sampling time and the previous load switch event is larger than the preset time threshold value, judging that the load switch event occurs and extracting corresponding load waveform data; extracting corresponding electric power fingerprint characteristics according to the extracted load waveform data corresponding to each load switch event, inputting the electric power fingerprint characteristics into a trained electric power fingerprint identification model, and outputting a load type identification result; the invention carries out data sampling and event detection based on the sliding window with the multistage edge distance windows at two ends, can effectively solve false detection caused by long transient state of load, fluctuation of abnormal values and surge current, and effectively improves the identification accuracy of electric fingerprint identification.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without inventive exercise.
Fig. 1 is a flowchart of a method for power fingerprinting based on event detection according to an alternative embodiment of the present invention;
FIG. 2 is a flowchart of a method for event detection based power fingerprinting in accordance with another alternative embodiment of the present invention;
FIG. 3 is a block diagram illustrating a structural connection of an event detection-based power fingerprinting device according to an alternative embodiment of the present invention;
fig. 4 is a block diagram illustrating a structural connection of an event detection-based power fingerprinting apparatus according to another alternative embodiment of the present invention.
Reference numerals:
1-a sampling module; 2-an event detection module; 3-an electric power fingerprint identification module; 4-an initial training set construction module; 5-training set enhancement module; 6-model training module.
Detailed Description
The embodiment of the invention provides an event detection-based power fingerprint identification method, device and equipment, which are used for solving the technical problem that the accurate identification rate of the existing event detection-based power fingerprint identification method still needs to be improved.
In order to make the objects, features and advantages of the present invention more obvious and understandable, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention provides an electric power fingerprint identification method based on event detection.
Referring to fig. 1, fig. 1 is a flowchart illustrating a method for power fingerprint identification based on event detection according to an embodiment of the present invention.
The embodiment of the invention provides an electric power fingerprint identification method based on event detection, which comprises the following steps:
the method comprises the following steps that S1, sliding windows with multi-stage edge distance windows at two ends are adopted to sample electrical quantity data; the length of the sliding window is greater than the sum of the lengths of the multistage edge distance windows at the two ends of the sliding window, and the multistage edge distance window at each end is composed of a main edge distance window and a secondary edge distance window.
In this embodiment, a sliding window with multi-level edge distance windows at both ends is used for data sampling. Specifically, let the length of the sliding window be N, and the length of the multi-stage edge distance window be N m The length of the main edge distance window is N m1 The length of the sub-edge distance window is N m2 Then each parameter satisfies 2N m ≤N,N m1 +N m2 ＝N m 。
And S2, after each sampling, respectively calculating the average values of the sampled electrical quantities of the main edge distance window at the left end of the sliding window, the secondary edge distance window at the left end, the main edge distance window at the right end and the secondary edge distance window at the right end, and if the absolute values of the difference values of the electrical quantities corresponding to the main edge distance windows at the two ends of the sliding window and the secondary edge distance windows are all larger than a preset electrical quantity threshold value, and the time interval between the current sampling time and the previous load switch event is larger than the preset time threshold value, judging that the load switch event occurs and extracting corresponding load waveform data.
In one implementation, the electrical quantity data is an active power sequence composed of active power of each cycle, and the active power is calculated by voltage and current measured by the corresponding cycle. Based on this, the specific process of determining the occurrence of the load switch event in step S2 is as follows: and if the absolute values of the difference values of the active power mean values corresponding to the main edge distance window and the secondary edge distance window at the two ends of the sliding window are both larger than a preset active power threshold value, and the time interval between the current sampling time and the previous load switch event is larger than a preset time threshold value, judging that the load switch event occurs.
As a specific embodiment, the calculation formula for determining whether a load switch event occurs is as follows:
in the formula, p j Is the active power of the jth cycle,
is j k The voltage sample value at the time of day,
is j is k Current sampling value at the moment, M is the number of sampling points of each cycle voltage/current, l and r are indexes of left and right boundary points of a sliding window at the current moment respectively, and N is m1 Is the length of the left/right main edge distance window, N m2 Is the length of the left/right minor edge from the window, P 1L 、P 1R 、P 2L 、P 2R Respectively corresponding active power mean values, delta P, of the left main margin window, the right main margin window, the left secondary margin window and the right secondary margin window 1 、ΔP 2 Respectively the difference value of the active power mean values corresponding to the main edge distance window and the secondary edge distance window, h is the preset active power threshold value of the occurrence of a load switch event, t lim For a preset time threshold, Δ t is the time interval from the current sampling time to the previous load switch event, event _ type = on indicates that a load on event occurs, event _ type = off indicates that a load off event occurs, and event _ type = none indicates that a load switch event is not transmitted.
The occurrence of a load switch event indicates the occurrence of a load attach event or the occurrence of a load remove event. In this embodiment, by setting a criterion formula, further determination of a load access event and a load shedding event is realized.
In one embodiment, the voltage/current sampling value is sampled by a smart socket. A suitable sampling frequency may be set, for example, a sampling frequency greater than or equal to 1000Hz.
In another possible implementation, the electrical quantity data may also be a sequence of current effective values calculated from current sample values of each period. Based on this, the specific process of determining the occurrence of the load switch event in step S2 is: and respectively calculating the current effective value mean values in the main edge distance window at the left end of the sliding window, the secondary edge distance window at the left end, the main edge distance window at the right end and the secondary edge distance window at the right end, and judging that a load switching event occurs if the absolute values of the difference values of the current effective value mean values corresponding to the main edge distance window and the secondary edge distance window at the two ends of the sliding window are all larger than a preset current effective value threshold value, and the time interval between the current sampling time and the previous load switching event is larger than a preset time threshold value.
According to the embodiment of the invention, an event detection scheme based on the multistage margin sliding window with time intervals is adopted, so that the effective detection of the load switch event is realized, the time intervals can be set to solve the problem of event false detection caused by long transient state when some loads are connected, the multistage margins can be set to solve the problems of abnormal value fluctuation of load data during measurement and event false detection caused by surge current of the loads, and the accuracy of power fingerprint identification can be effectively improved on the whole.
And S3, extracting corresponding power fingerprint characteristics according to the extracted load waveform data corresponding to each load switch event, inputting the power fingerprint characteristics into a trained power fingerprint identification model, and outputting a load type identification result.
In one implementation, the corresponding load waveform data includes voltage and current data before and after the occurrence of the corresponding load switching event, and the extracting the corresponding power fingerprint feature according to the extracted load waveform data corresponding to each load switching event includes:
segmenting voltage and current data before and after each load switch event by taking a zero crossing point of a voltage rising stage as a boundary, so that each obtained data segment is data of one cycle;
if the load switch event is the load access, subtracting the current data before the corresponding event from the current data section after the corresponding event occurs to obtain separated current data; if the load switch event is load cutting, subtracting the current data after the corresponding event from the current data before the corresponding event to obtain separated current data;
merging the separated current data with the voltage data section after the corresponding event to obtain separated load data;
and extracting the electric power fingerprint characteristics of each separated load data.
In this embodiment, the voltage and current data before and after each load switch event is segmented by using the zero-crossing point of the voltage rising stage as a boundary, so that waveform alignment can be realized, and thus, the load can be conveniently and directly separated by matrix operation in the subsequent process, and the separated load data can be obtained. The segment data U of the voltage and current of j period obtained by the step j 、I j Can be expressed as:
in the formula (I), the compound is shown in the specification,
respectively represent the j th 1 ,j 2 ,…,j k ,…,j M The voltage sampling value at a moment, M is the number of sampling points of voltage/current of each cycle (namely, period),
respectively represent the j th 1 ,j 2 ,…,j k ,…,j M Current sample values at time instants.
The expression for separating the loads by matrix operation can be expressed as:
j_before_event＜j_event_flag
j_after_event＞j_event_flag
in the formula I dec Representing post-separation current data, I j_after_event Representing a current data section after an event has occurred, I j_before_event The current data segment before the event occurs is shown, j _ event _ flag is shown as the cycle number when the event occurs, j _ before _ event is shown as the cycle number before the event occurs, and j _ after _ event is shown as the cycle number after the event occurs.
Further, the expression of the load data after separation is:
X dec ＝{U j_after_event ,I dec }
in the formula, X dec Representing load data after separation, U j_after_event Representing the voltage data segment after the corresponding event.
In one implementation, the extracting the power fingerprint feature of each separated load data includes:
extracting power characteristics, harmonic characteristics and waveform characteristics of the separated load data; the power characteristics include active power, reactive power, apparent power and power factor; the harmonic characteristics comprise effective values and phases of fundamental wave, 2, 3, 5 and 7 times of harmonic current and impedance, and further comprise 4-6-8-10 times of current harmonic aggregation effective values, 9-12-15 times of current harmonic aggregation effective values and 11-13-17-19-23-25 times of current harmonic aggregation effective values; the waveform characteristics include a total harmonic distortion rate, a form factor, and a crest factor.
Wherein, the effective value and the phase of each harmonic can be obtained by Fourier transform calculation.
In one implementation, the current harmonic aggregation effective value is calculated as follows:
in the formula I Σ Represents the effective value of current harmonic aggregation, I h Representing the effective value of the H harmonic current, H being the set of harmonic orders aggregated.
In an enabling manner, the total harmonic distortion rate THD i The calculation formula of (a) is as follows:
in the formula I 1 Representing the effective value of the fundamental current.
In one implementation, the form factor k is a The calculation formula of (a) is as follows:
in the formula I arv Is the absolute mean value of the current, I rms Is the effective value of the current i k M is the current value number in one cycle for the kth current value.
In one implementation, the crest factor k f Is calculated as followsThe following steps:
in the formula I peak Is the peak current value.
According to the embodiment of the invention, the power fingerprint characteristics comprise power characteristics, harmonic characteristics and waveform characteristics, the connotation of load characteristics is further expanded on the basis of the prior art, the power utilization state of equipment can be more accurately expressed, and more precise identification is realized.
Referring to fig. 2, fig. 2 is a flowchart illustrating a power fingerprinting method based on event detection according to another alternative embodiment of the present invention.
In an implementation manner, on the basis of the method shown in fig. 1, the method described in this embodiment further includes:
s4, collecting and labeling the steady-state load waveform data of each type of single load, extracting the power fingerprint characteristics of the corresponding single load according to the labeled steady-state load waveform data, and constructing an initial training set;
s5, performing data enhancement on the initial training set to obtain a training set after data enhancement;
and S6, training the electric power fingerprint identification model by using the training set after the data enhancement to obtain the trained electric power fingerprint identification model.
Aiming at the problems of fewer load samples and higher data collection cost in the prior art, in the embodiment of the invention, the data enhancement is carried out on the initial training set, and the training set after the data enhancement is used for training the electric fingerprint identification model, so that the robustness of the model can be effectively improved under the conditions of fewer load samples and higher data collection cost, and the identification accuracy of the load under the disturbance of the combined load is improved.
In one implementation, the extracting the power fingerprint feature of the corresponding individual load according to the tagged steady-state load waveform data includes:
preprocessing the labeled steady-state load waveform data; the preprocessing comprises filling data missing values, removing abnormal data, removing data which are not matched with the corresponding labels and/or removing data which are not matched with the corresponding switch states;
and extracting the electric power fingerprint characteristics of the corresponding single load from the preprocessed steady-state load waveform data.
In this embodiment, the precision of the data enhancement operation is improved by preprocessing the labeled steady-state load waveform data.
In one implementation, the data enhancing the initial training set includes:
selecting a group of data from the preprocessed steady-state load waveform data, segmenting the data according to a zero crossing point of a voltage rising stage so as to enable each obtained data segment to be data of a cycle wave, and selecting two sections of unequal current waveforms from each obtained data segment for subtraction to obtain corresponding disturbance waveform data;
repeating the previous step to obtain a plurality of disturbance waveform data, and constructing a disturbance waveform data set according to the plurality of disturbance waveform data;
selecting any group of data from the preprocessed steady-state load waveform data, segmenting the data according to a zero crossing point of a voltage rising stage, determining a target current data segment from each obtained data segment, randomly selecting a plurality of disturbance waveform data from the disturbance waveform data set to be superposed on the target current data segment, and obtaining reconstructed load data without changing voltage data and labels corresponding to the target current data segment;
repeating the previous step to obtain a plurality of reconstruction load data, and constructing a reconstruction load data set according to the plurality of reconstruction load data;
and extracting the electric power fingerprint characteristics of the corresponding individual loads from each reconstructed load data in the reconstructed load data set, and adding the electric power fingerprint characteristics into the initial training set.
In the embodiment of the invention, the process of generating the reconstructed load data is basically consistent with the process of separating the target load data from the combined load, the target load data under the combined load disturbance can be simulated, the more essential fingerprint characteristics of the load can be searched by the model under the influence of the combined load disturbance, and the identification accuracy of the load under the disturbance of the combined load is improved.
In one implementation, the training of the power fingerprinting model using the data-enhanced training set includes:
and constructing a random forest model as the electric power fingerprint identification model.
The random forest is a supervised integrated learning classification technology, a classification model of the random forest is composed of a group of decision tree classifiers, and the classification of data by the model is to determine a final result by collectively voting through classification results of single decision trees. The method combines the Bagging integrated learning theory of Leo Breiman with the random subspace method proposed by Ho, fully ensures the independence and difference between each decision tree by injecting randomness into the training sample space and the attribute space, well overcomes the problem of over-fitting of the decision trees, and has better robustness to noise and abnormal values. In the embodiment, the random forest model is used as the electric power fingerprint identification model, so that certain robustness in the electric power fingerprint identification process can be guaranteed.
It should be noted that, in other embodiments, other suitable machine learning models may also be used as the power fingerprint identification model.
The invention also provides an electric power fingerprint identification device based on event detection, which can be used for executing the electric power fingerprint identification method based on event detection in any one of the above embodiments of the invention.
Referring to fig. 3, fig. 3 is a block diagram illustrating a structural connection of an electric fingerprint identification device based on event detection according to an embodiment of the present invention.
The embodiment of the invention provides an electric power fingerprint identification device based on event detection, which comprises:
the sampling module 1 is used for sampling the electrical quantity data by adopting a sliding window with multistage edge distance windows arranged at two ends; the length of the sliding window is greater than the sum of the lengths of the multistage edge distance windows at the two ends of the sliding window, and the multistage edge distance window at each end consists of a main edge distance window and a secondary edge distance window;
the event detection module 2 is used for respectively calculating the average values of the sampled electrical quantities of the main edge distance window at the left end of the sliding window, the secondary edge distance window at the left end, the main edge distance window at the right end and the secondary edge distance window at the right end after sampling each time, and if the absolute values of the difference values of the electrical quantity average values corresponding to the main edge distance windows and the secondary edge distance windows at the two ends of the sliding window are all larger than a preset electrical quantity threshold value, and the time interval between the current sampling time and the previous load switch event is larger than a preset time threshold value, judging that the load switch event occurs and extracting corresponding load waveform data;
and the electric power fingerprint identification module 3 is used for extracting corresponding electric power fingerprint characteristics according to the extracted load waveform data corresponding to each load switch event, inputting the electric power fingerprint characteristics into a trained electric power fingerprint identification model, and outputting a load type identification result.
In one implementation, the electrical quantity data is an active power sequence composed of active power of each cycle, and the active power is calculated by voltage and current measured by the corresponding cycle.
In one implementation, the corresponding load waveform data includes voltage and current data before and after a corresponding load switch event occurs, and the power fingerprinting module 3 includes:
the waveform alignment unit is used for segmenting voltage and current data before and after each load switch event by taking a zero crossing point of a voltage rising stage as a boundary so as to enable each obtained data segment to be data of one cycle;
the first increment extraction unit is used for subtracting the current data before the corresponding event from the current data section after the corresponding event occurs to obtain separated current data if the load switch event is the load access; if the load switch event is load cutting, subtracting the current data after the corresponding event from the current data section before the corresponding event to obtain separated current data;
the data merging unit is used for merging the separated current data with the voltage data section after the corresponding event to obtain separated load data;
and the first characteristic extraction unit is used for extracting the electric power fingerprint characteristics of each separated load data.
In one possible implementation manner, the first feature extraction unit includes:
the characteristic extraction subunit is used for extracting the power characteristic, the harmonic characteristic and the waveform characteristic of the separated load data; the power characteristics include active power, reactive power, apparent power and power factor; the harmonic characteristics comprise effective values and phases of fundamental wave, 2, 3, 5 and 7 times of harmonic current and impedance, and further comprise 4-6-8-10 times of current harmonic aggregation effective values, 9-12-15 times of current harmonic aggregation effective values and 11-13-17-19-23-25 times of current harmonic aggregation effective values; the waveform characteristics include a total harmonic distortion rate, a form factor, and a crest factor.
In an implementation manner, the feature extraction subunit is specifically configured to:
calculating the polymerization effective value of each current harmonic according to the following formula:
in the formula I Σ Represents the effective value of current harmonic aggregation, I h Representing the effective value of H harmonic current, H being the set of harmonic orders aggregated.
Referring to fig. 4, fig. 4 is a block diagram illustrating a structural connection of an electric fingerprint identification device based on event detection according to an embodiment of the present invention.
In one implementation manner, as shown in fig. 4, based on the structure of the apparatus shown in fig. 3, the apparatus further includes:
the initial training set building module 4 is used for collecting the steady-state load waveform data of each type of single load, labeling the steady-state load waveform data, extracting the power fingerprint characteristics of the corresponding single load according to the labeled steady-state load waveform data, and building an initial training set;
a training set enhancing module 5, configured to perform data enhancement on the initial training set to obtain a training set after data enhancement;
and the model training module 6 is used for training the electric power fingerprint identification model by using the training set after the data enhancement to obtain the trained electric power fingerprint identification model.
In one implementation, the initial training set building module 4 includes:
the data preprocessing unit is used for preprocessing the labeled steady-state load waveform data; the preprocessing comprises filling data missing values, removing abnormal data, removing data which are not matched with the corresponding labels and/or removing data which are not matched with the corresponding switch states;
and the second characteristic extraction unit is used for extracting the electric power fingerprint characteristics of the corresponding single load from the preprocessed steady-state load waveform data.
In one implementation, the training set enhancing module 5 includes:
the second increment extraction unit is used for selecting a group of data from the preprocessed steady-state load waveform data, segmenting the data according to the zero crossing point of the voltage rising stage so as to enable each obtained data segment to be data of a cycle wave, and selecting two unequal current waveforms from each obtained data segment for subtraction to obtain corresponding disturbance waveform data;
the disturbance waveform data set construction unit is used for repeating the second increment extraction unit to obtain a plurality of disturbance waveform data and constructing a disturbance waveform data set according to the plurality of disturbance waveform data;
a third increment extraction unit, configured to select any group of data from the preprocessed steady-state load waveform data, segment the data according to a zero crossing point of a voltage rising stage, determine a target current data segment from each obtained data segment, randomly select a plurality of disturbance waveform data from the disturbance waveform data set, superimpose the disturbance waveform data onto the target current data segment, and obtain reconstructed load data without changing voltage data and a label corresponding to the target current data segment;
the reconstructed load data set construction unit is used for repeating the third increment extraction unit to obtain a plurality of reconstructed load data and constructing a reconstructed load data set according to the plurality of reconstructed load data;
and the third feature extraction unit is used for extracting the electric power fingerprint features of the corresponding individual loads from each reconstructed load data in the reconstructed load data set and adding the electric power fingerprint features into the initial training set.
In one implementation, the model training module 6 includes:
and the model construction unit is used for constructing a random forest model as the electric power fingerprint identification model.
The invention also provides an electric power fingerprint identification device based on event detection, which comprises:
a memory to store instructions; the instructions are used for realizing the power fingerprint identification method based on event detection as described in any one of the above embodiments;
a processor to execute the instructions in the memory.
The present invention also provides a computer-readable storage medium, having a computer program stored thereon, where the computer program, when executed by a processor, implements the power fingerprint identification method based on event detection as described in any one of the above embodiments.
It can be clearly understood by those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described apparatuses, devices and modules may refer to the corresponding processes in the foregoing method embodiments, and the specific beneficial effects of the above-described apparatuses, devices and modules may refer to the corresponding beneficial effects in the foregoing method embodiments, which are not described herein again.
In the several embodiments provided in the present application, it should be understood that the disclosed apparatus, device and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the modules is only one logical division, and other divisions may be realized in practice, for example, a plurality of modules or components may be combined or integrated into another apparatus, or some features may be omitted, or not executed.
The modules described as separate parts may or may not be physically separate, and parts displayed as modules may or may not be physical modules, may be located in one position, or may be distributed on a plurality of network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, functional modules in the embodiments of the present invention may be integrated into one processing module, or each of the modules may exist alone physically, or two or more modules are integrated into one module. The integrated module can be realized in a hardware mode, and can also be realized in a software functional module mode.
The integrated module, if implemented in the form of a software functional module and sold or used as a separate product, may be stored in a computer-readable storage medium. Based on such understanding, the technical solution of the present invention, which is substantially or partly contributed by the prior art, or all or part of the technical solution may be embodied in a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
The above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; although the present invention has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (12)
1. A power fingerprint identification method based on event detection is characterized by comprising the following steps:
sampling electrical quantity data by adopting a sliding window with multistage edge distance windows arranged at two ends; the length of the sliding window is greater than the sum of the lengths of the multistage edge distance windows at the two ends of the sliding window, and the multistage edge distance window at each end consists of a main edge distance window and a secondary edge distance window;
after sampling each time, respectively calculating the average values of the sampled electrical quantities of the main edge distance window at the left end of the sliding window, the secondary edge distance window at the left end, the main edge distance window at the right end and the secondary edge distance window at the right end, and if the absolute values of the difference values of the electric quantity average values corresponding to the main edge distance windows and the secondary edge distance windows at the two ends of the sliding window are all larger than a preset electrical quantity threshold value and the time interval from the current sampling time to the previous load switch event is larger than a preset time threshold value, judging that the load switch event occurs and extracting corresponding load waveform data;
and extracting corresponding power fingerprint characteristics according to the extracted load waveform data corresponding to each load switch event, inputting the power fingerprint characteristics into a trained power fingerprint identification model, and outputting a load type identification result.
2. The event detection-based power fingerprint identification method according to claim 1, wherein the electrical quantity data is an active power sequence consisting of active power of each period, and the active power is calculated by voltage and current measured by the corresponding period.
3. The method according to claim 1, wherein the corresponding load waveform data includes voltage and current data before and after the occurrence of the corresponding load switch event, and the extracting the corresponding power fingerprint feature according to the extracted load waveform data corresponding to each load switch event includes:
segmenting voltage and current data before and after each load switch event by taking a zero crossing point of a voltage rising stage as a boundary, so that each obtained data segment is data of one cycle;
if the load switch event is the load access, subtracting the current data before the corresponding event from the current data section after the corresponding event occurs to obtain separated current data; if the load switch event is load cutting, subtracting the current data after the corresponding event from the current data section before the corresponding event to obtain separated current data;
merging the separated current data with the voltage data section after the corresponding event to obtain separated load data;
and extracting the electric power fingerprint characteristics of each separated load data.
4. The event detection-based power fingerprint identification method according to claim 3, wherein the extracting the power fingerprint feature of each separated load data comprises:
extracting power characteristics, harmonic characteristics and waveform characteristics of the separated load data; the power characteristics include active power, reactive power, apparent power and power factor; the harmonic characteristics comprise effective values and phases of fundamental wave, 2, 3, 5 and 7 times of harmonic current and impedance, and further comprise 4-6-8-10 times of current harmonic aggregation effective values, 9-12-15 times of current harmonic aggregation effective values and 11-13-17-19-23-25 times of current harmonic aggregation effective values; the waveform characteristics include a total harmonic distortion rate, a form factor, and a crest factor.
5. The event detection-based power fingerprinting method of claim 4, characterized in that the extracting of the power, harmonic and waveform characteristics of the separated load data comprises:
calculating the harmonic aggregation effective value of each current according to the following formula:
in the formula I Σ Represents the effective value of current harmonic aggregation, I h Representing the effective value of H harmonic current, H being the set of harmonic orders aggregated.
6. The event detection-based power fingerprinting method of claim 1, characterized in that it further comprises:
collecting and labeling steady state load waveform data of each type of single load, extracting power fingerprint characteristics of corresponding single loads according to the labeled steady state load waveform data, and constructing an initial training set;
performing data enhancement on the initial training set to obtain a training set after data enhancement;
and training the electric power fingerprint identification model by using the training set after the data enhancement to obtain the trained electric power fingerprint identification model.
7. The method according to claim 6, wherein the extracting the power fingerprint feature of the corresponding individual load according to the tagged steady-state load waveform data comprises:
preprocessing the labeled steady-state load waveform data; the preprocessing comprises filling data missing values, removing abnormal data, removing data which are not matched with the corresponding labels and/or removing data which are not matched with the corresponding switch states;
and extracting the electric power fingerprint characteristics of the corresponding single load from the preprocessed steady-state load waveform data.
8. The event detection-based power fingerprinting method of claim 7, wherein the data enhancement of the initial training set comprises:
selecting a group of data from the preprocessed steady-state load waveform data, segmenting the data according to a zero crossing point of a voltage rising stage so as to enable each obtained data segment to be data of a cycle wave, and selecting two sections of unequal current waveforms from each obtained data segment for subtraction to obtain corresponding disturbance waveform data;
repeating the previous step to obtain a plurality of disturbance waveform data, and constructing a disturbance waveform data set according to the plurality of disturbance waveform data;
selecting any group of data from the preprocessed steady-state load waveform data, segmenting the data according to a zero crossing point of a voltage rising stage, determining a target current data segment from each obtained data segment, randomly selecting a plurality of disturbance waveform data from the disturbance waveform data set to be superposed on the target current data segment, and obtaining reconstructed load data without changing voltage data and labels corresponding to the target current data segment;
repeating the previous step to obtain a plurality of reconstruction load data, and constructing a reconstruction load data set according to the plurality of reconstruction load data;
and extracting the electric power fingerprint characteristics of the corresponding individual load from each reconstructed load data in the reconstructed load data set, and adding the electric power fingerprint characteristics into the initial training set.
9. The method according to claim 6, wherein training a power fingerprinting model using the training set enhanced by the data comprises:
and constructing a random forest model as the electric power fingerprint identification model.
10. An event detection-based power fingerprint identification device, comprising:
the sampling module is used for sampling the electrical quantity data by adopting a sliding window with multistage edge distance windows arranged at two ends; the length of the sliding window is greater than the sum of the lengths of the multistage edge distance windows at the two ends of the sliding window, and the multistage edge distance window at each end consists of a main edge distance window and a secondary edge distance window;
the event detection module is used for respectively calculating the average values of the sampled electrical quantities of the main edge distance window at the left end of the sliding window, the secondary edge distance window at the left end, the main edge distance window at the right end and the secondary edge distance window at the right end after sampling each time, and if the absolute values of the difference values of the electrical quantities corresponding to the main edge distance windows at the two ends of the sliding window and the secondary edge distance windows are all larger than a preset electrical quantity threshold value, and the time interval between the current sampling time and the previous load switch event is larger than the preset time threshold value, judging that the load switch event occurs and extracting corresponding load waveform data;
and the electric power fingerprint identification module is used for extracting corresponding electric power fingerprint characteristics according to the extracted load waveform data corresponding to each load switch event, inputting the electric power fingerprint characteristics into a trained electric power fingerprint identification model and outputting a load type identification result.
11. An event detection-based power fingerprinting device, characterized by comprising:
a memory to store instructions; wherein the instructions are used for realizing the power fingerprint identification method based on the event detection according to any one of claims 1 to 9;
a processor to execute the instructions in the memory.
12. A computer-readable storage medium, having stored thereon a computer program which, when executed by a processor, implements the event detection-based power fingerprinting method according to any of the claims 1-9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211175286.8A CN115564324A (en) | 2022-09-26 | 2022-09-26 | Electric power fingerprint identification method, device and equipment based on event detection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211175286.8A CN115564324A (en) | 2022-09-26 | 2022-09-26 | Electric power fingerprint identification method, device and equipment based on event detection |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115564324A true CN115564324A (en) | 2023-01-03 |
Family
ID=84742328
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211175286.8A Pending CN115564324A (en) | 2022-09-26 | 2022-09-26 | Electric power fingerprint identification method, device and equipment based on event detection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115564324A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115907567A (en) * | 2023-02-21 | 2023-04-04 | 浙江大学 | Load event detection method and system based on robustness random segmentation forest algorithm |
-
2022
- 2022-09-26 CN CN202211175286.8A patent/CN115564324A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115907567A (en) * | 2023-02-21 | 2023-04-04 | 浙江大学 | Load event detection method and system based on robustness random segmentation forest algorithm |
| CN115907567B (en) * | 2023-02-21 | 2023-05-09 | 浙江大学 | Load event detection method and system based on robust random segmentation forest algorithm |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Mahela et al. | Recognition of power quality disturbances using S-transform based ruled decision tree and fuzzy C-means clustering classifiers | |
| CN104463706B (en) | Method and system for detecting voltage sag event reason for power grid | |
| CN110108992B (en) | Cable partial discharge fault identification method and system based on improved random forest algorithm | |
| Zhu et al. | Spatial–temporal feature learning in smart grids: A case study on short-term voltage stability assessment | |
| Shi et al. | Fault classification for transmission lines based on group sparse representation | |
| CN109800660B (en) | Voltage sag source identification method and system based on big data clustering | |
| CN112750051B (en) | Random forest algorithm-based phase sequence identification method and device for transformer area and terminal equipment | |
| CN112085111B (en) | Load identification method and device | |
| CN111537853A (en) | Intelligent detection method for partial discharge of switch cabinet based on multi-source heterogeneous data analysis | |
| CN115564324A (en) | Electric power fingerprint identification method, device and equipment based on event detection | |
| CN114325256A (en) | Power equipment partial discharge identification method, system, equipment and storage medium | |
| CN116679161A (en) | Power grid line fault diagnosis method, equipment and medium | |
| Asman et al. | Identification of transient overvoltage using discrete wavelet transform with minimised border distortion effect and support vector machine | |
| CN111898694B (en) | Non-invasive load identification method and device based on random tree classification | |
| CN111896890B (en) | Micro-grid line fault diagnosis method and system based on Hilbert-Huang transform | |
| CN113036754A (en) | Non-invasive load identification method based on improved DAG-SVMS | |
| CN115186772B (en) | Method, device and equipment for detecting partial discharge of power equipment | |
| CN111898446A (en) | Single-phase earth fault studying and judging method based on multi-algorithm normalization analysis | |
| CN114764599B (en) | Power distribution network single-phase earth fault sensitivity analysis method and system | |
| CN116365511A (en) | Active power distribution network model construction method, device, terminal and storage medium | |
| CN116127447A (en) | Virtual power plant false data injection attack detection method, device, terminal and medium | |
| CN109670526A (en) | A kind of interference source type discrimination method, device, equipment and the storage medium of power grid | |
| CN112731208B (en) | Low-voltage line fault and abnormity on-line monitoring method, equipment and medium | |
| Li et al. | Fault diagnosis for single-phase grounding in distribution network based on hilbert-huang transform and siamese convolution neural network | |
| Zhu et al. | Fault Phase Identification Method Based on Convolutional Neural Network for Double Circuit Transmission Lines |
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 |