CN115330202A - Data-driven low-voltage distribution station area electricity stealing analysis method - Google Patents
Data-driven low-voltage distribution station area electricity stealing analysis method Download PDFInfo
- Publication number
- CN115330202A CN115330202A CN202210976257.5A CN202210976257A CN115330202A CN 115330202 A CN115330202 A CN 115330202A CN 202210976257 A CN202210976257 A CN 202210976257A CN 115330202 A CN115330202 A CN 115330202A
- Authority
- CN
- China
- Prior art keywords
- data
- line loss
- electricity stealing
- electricity
- analyzed
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/02—Knowledge representation; Symbolic representation
- G06N5/022—Knowledge engineering; Knowledge acquisition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Abstract
The invention discloses a data-driven low-voltage distribution station area electricity stealing analysis method, which comprises the following steps: calculating the data characteristics of the to-be-analyzed region; counting similar transformer areas of the transformer area to be analyzed, constructing a line loss identification model, and decomposing line loss characteristics of the transformer area to be analyzed; constructing a knowledge base and defining logic derivation rules of the knowledge base; and combining the line loss characteristics and the data characteristics of the users in the distribution area to be analyzed to form a sample vector group, and performing electricity stealing analysis by using the knowledge base. The method is used for analyzing the power stealing of the distribution room based on data driving, does not need a large number of labeled power stealing samples for training, has wider use value, and is suitable for analyzing the power stealing of the low-voltage distribution room in the complex line loss environment.
Description
Technical Field
The invention relates to the technical field of power monitoring, in particular to a data-driven low-voltage distribution station area electricity stealing analysis method.
Background
At present, the common method for realizing the electricity stealing analysis of the power distribution station area comprises the following steps: firstly, a traditional machine learning-based electricity stealing analysis method is generally a multi-classification model or a logistic regression model, an electricity stealing data sample is collected to train an electricity stealing research and judgment model, the method belongs to a classic machine learning algorithm, the fitting performance and the running speed are well represented, but the method needs a label value of the sample, namely whether the sample is an electricity stealing user or not, electricity stealing information is sensitive, and the quantity/quality of data is high; secondly, a power stealing analysis method based on a constraint least square method does not need sample label support, the method maps a power stealing rule into a coefficient of a regression algorithm, the constraint coefficient and the contraction coefficient are in a specified range, the power stealing of a user is judged through the analysis coefficient, and the common methods are inhaul cable regression and ridge regression, but the method is weak in applicability and not suitable for the conditions that the correlation between the electric quantity of the user and the line loss of a transformer area is weak, and the relative line loss average value of a calculation period fluctuates too much (such as the conditions of flying line in front of a meter, intermittent power stealing and the like); thirdly, a branch monitoring unit is additionally arranged, the electric quantity of a line and the electric quantity of a terminal metering device (an intelligent switch and an electric meter) are collected through the branch monitoring unit, comparison and study are carried out, the method belongs to electricity stealing judgment realized by hardware, is accurate and short in calculation period, but relates to manual point selection, installation, testing and other work, hardware and labor cost are high, and popularization conditions cannot be met in a low-voltage power distribution station area.
Disclosure of Invention
The invention provides a data-driven low-voltage distribution station area electricity stealing analysis method, which aims to: the method overcomes the defects of the prior art, and provides the electricity stealing analysis method with strong applicability on the premise of not needing a large amount of sample label support and not increasing hardware cost.
The technical scheme of the invention is as follows:
a data-driven low-voltage distribution station area electricity stealing analysis method comprises the following steps:
s1: calculating the zone Z to be analyzed * The data characteristic of (a);
s2: statistics of the region Z to be analyzed * Similar station zone Z of i Constructing a line loss identification model, and decomposing line loss characteristics of the transformer area to be analyzed;
s3: constructing a knowledge base, and defining logic derivation rules of the knowledge base;
s4: to-be-analyzed station region Z * And combining the line loss characteristics and the data characteristics of the lower user to form a sample vector group, and performing electricity stealing analysis by using the knowledge base.
Further, in step S2, the method for constructing the line loss identification model includes:
s21: line loss characteristics of all electricity stealing behavior types are abstracted into the following three types:
i: the electricity stealing users have a correlation between electricity and line loss electricity;
j: the electricity stealing users have no correlation between electricity and line loss electricity;
k: the line loss electric quantity is close to a constant value;
the line loss sequence expression is as follows:
I＝{a 1 (X 1 ) 1 +…+a n (X n ) 1 ,...,a 1 (X 1 ) m +…+a n (X n ) m }
J＝{b 1 ,b 2 ,...,b m }
wherein (X) n ) m For the zone Z to be analyzed * The electric quantity value of the electric meter is I, the electric quantity of the electric meter is multiplied by the addition sequence of the coefficients, J is an independent power consumption independent sequence, and K is an electric quantity constant sequence;
s22: constructing a transformer area line loss model:
a j ≥0,b j ≥0,c≥0
wherein eta 1 、η 2 、η 3 Is a hyper-parameter and is used for restricting the space of a search solution, m is the moment of electric quantity data, D represents a station area Z to be analyzed * Line loss of d LIP To move the multiline position distance.
J = { b = { [ b ] 1 ,b 2 ,...,b m Divide it into sub subsequences J sub A mixture of J and sub similar station zone Z i The moving multiline position distance is calculated by the curve sequence of the electric quantity in (1), d LIP The calculation formula is as follows:
wherein, t p Is J sub And region Z i Intersection of medium electric quantity curve sequences, area p Is J sub And region Z i The area of a polygonal area formed by the medium electric quantity curves.
In turn with Z i Each electric quantity on the curve sequence of the medium electric quantitySliding a certain J in the vertical and horizontal directions by using the point corresponding to the value as a reference sub Sequence, and calculating J sub And Z i The minimum distance matched in the sliding process is recorded.
All J are put together sub And summing the matched minimum distances to serve as a constraint term of the line loss model.
Further, in step S2, the method for decomposing the line loss characteristics of the to-be-analyzed distribution room includes: and (3) performing solution space search on minLoss (a, b and c) by using a meta-heuristic algorithm, and judging the numerical values of the line loss characteristics of I, J and K when the algorithm converges to obtain the line loss characteristics of the transformer area.
Further, the method for constructing the knowledge base in step S3 includes: the method comprises the following steps of counting the data characteristics of known electricity stealing users and suspected electricity stealing users under a system, performing knowledge creation by combining power service knowledge, establishing the relation between the data characteristics and line loss characteristics to electricity stealing events, and forming a knowledge base, wherein the association of the line loss characteristics in the knowledge base is performed according to the following principle:
A. when the line loss electricity quantity and the user electricity consumption quantity are generated by the same group of electric equipment in the same period, the line loss characteristic is I;
B. when the line loss electricity is generated by completely independent electric equipment, the line loss characteristic is J;
C. when the line loss electric quantity can be manually controlled or intervened, the line loss characteristic is K;
and if one electricity stealing event can meet a plurality of the A, the B and the C, forming a plurality of corresponding knowledge of the electricity stealing event.
The logic derivation rule of the knowledge base comprises the following three results when a user steals the electricity sample is judged:
firstly, the electricity stealing sample certainly belongs to a knowledge base set, namely, the line loss characteristics of the electricity stealing sample can be formed by one or more pieces of knowledge in the knowledge base, and the user is judged to be a suspected electricity stealing user;
secondly, the electricity stealing samples may or may not belong to a knowledge base set, the description is carried out through approximate knowledge, the matching degree is calculated according to the columns in the knowledge base, a plurality of knowledge combinations with the matching degree larger than a preset value are selected, and the possibility that the user is a suspected electricity stealing user is judged according to the electricity stealing event with dominant position in the knowledge combinations;
thirdly, the electricity stealing samples do not necessarily belong to the knowledge base set, and the user eliminates the possibility of electricity stealing.
Further, the logic derivation rule of the knowledge base in step S3 further includes an overlap operator defining the line loss characteristics of I, J, and K:
where n denotes that two data sequences occur simultaneously,
when I and J are superposed, a new uncorrelated relation J between electricity stealing users and line loss electricity is generated * ，
When the I and the K are superposed, a new relation I between the electricity stealing user electricity quantity and the line loss electricity quantity is generated * ，
The trend characteristics are not changed when J and K are superposed, and the superposition result is still J.
When a user electricity stealing sample is judged, if the line loss characteristics of the electricity stealing sample can be formed by combining a plurality of pieces of knowledge in a knowledge base, during superposition operation of the plurality of pieces of knowledge, the line loss characteristics are subjected to superposition operation according to the definition of a superposition operation operator, data characteristics are subjected to operation according to a large value domination principle, corresponding electricity stealing events are superposed to form an array set, and the user is judged to be a suspected electricity stealing user with various electricity stealing behaviors.
Further, the step S1 is to calculate the zone Z to be analyzed * Before the data characteristic of (2) the zone to be analyzed Z is first treated * And performing data filling processing by the following method:
s11: zone to be analyzed Z * The data matrix form of (a) is:
wherein (X) n ) m Representing the electric quantity of the electric meter, n is the electric meter identification under the transformer area, m is the moment of electric quantity data, and X 0 In particular to a table area general table;
longitudinal axis direction (X) 0 ) i (X 1 ) i …(X n ) i If the number of the missing data exceeds the set value, deleting the electric quantity data at the moment, and going to the step S12, otherwise, directly executing the step S12.
S12: for missing data in the matrix, the missing data is in the horizontal axis direction (X) corresponding to the missing data i ) 1 (X i ) 2 …(X i ) m The missing value is filled by polynomial interpolation, and the specific method is as follows:
the moment of the electric quantity data and the electric quantity value are regarded as point coordinates on a two-dimensional plane, the moment of the electric quantity data is an abscissa, the electric quantity value is an ordinate, and the electric quantity data at k moments are expressed as (x) 1 ,y 1 )(x 1 ,y 1 )…(x k ,y k ) And constructing a k-1 degree polynomial, and enabling known k coordinates to be established after being substituted into the k-1 degree polynomial, wherein the k-1 degree polynomial is as follows:
wherein L (x) is an interpolation expression, L j (x) Is an interpolation basis function, x is the value of data time, y is the value of electric quantity, and j represents the number of addition;
in [1, k ]]Introduces the (k + 1) th coordinate (x) in the range domain k+1 ,y k+1 ) Using said k-1 th order polynomial to carry the coordinates (x) k+1 ,y k+1 ) And solving the corresponding electric quantity value to complete the first-order polynomial interpolation filling.
S13: filling all missing data according to the method in the step S12 in sequence to finish the platform zone Z to be analyzed * The data of (3) is updated.
Further, the step S1 is to-be-analyzed station zone Z * When the data padding process is performed, k =24.
Further, step S2 is that the station area Z to be analyzed is counted * Similar station zone Z of i The method comprises the following steps:
extracting platform region vectors in batches from a service system, and calculating Z of each platform region and a platform region to be analyzed by using cosine similarity * The distance between:
wherein alpha represents a station zone Z * Beta represents a batch extracted station area vector in the service system, alpha i 、β i Respectively, the components in the vectors alpha, beta;
similarity calculation is carried out on alpha and a distribution area vector beta in a service system in sequence, and a distribution area Z to be analyzed is selected * Taking L nearest station areas as similar station areas Z i 。
Further, the step S2 further includes, according to the step S1, processing the zone Z to be analyzed * Method for filling data into similar region Z i And performing data filling processing.
Compared with the prior art, the invention has the following beneficial effects:
(1) The method is used for analyzing the electricity stealing of the transformer area based on data driving, the line loss characteristics of the transformer area are obtained by calculating the data characteristics of the transformer area to be analyzed and constructing a line loss identification model, a knowledge base and an inference method of the knowledge base are constructed to analyze the electricity stealing of the transformer area, a large number of labeled electricity stealing samples are not needed for training, the method has wider use value, and the method is suitable for analyzing the electricity stealing of the low-voltage power distribution transformer area in a complex line loss environment;
(2) The method provides the concept that the line loss electricity quantity can be abstracted into three types of line loss characteristics and line loss electricity quantity identification for the first time, and as a characteristic with higher importance, the characteristic dimension of line loss analysis is expanded from the data perspective, and the characteristic electricity stealing user is difficult to hide, so that the accuracy of line loss analysis is greatly improved;
(3) The method establishes the knowledge base, provides the characteristic of the electricity stealing users through the superposition operation of the knowledge base, defines the superposition operation mode of the data characteristic and the line loss characteristic, flexibly considers the situation that the same electricity stealing user may have various electricity stealing behaviors compared with a single type of electricity stealing method, and improves the applicability and the application range.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a diagram illustrating a method for calculating the characteristics of the data in the distribution room;
FIG. 3 is a diagram illustrating characteristics of subscriber data in a distribution area;
FIG. 4 is a schematic diagram of region coding;
FIG. 5 is an abstract view of line loss characteristics of electricity stealing behavior;
FIG. 6 is a schematic view of the multi-line position distance;
fig. 7 is a schematic diagram of subscriber line loss characteristics under a distribution area;
FIG. 8 is a schematic diagram of a knowledge base;
FIG. 9 is a schematic view of user sample vectors under a distribution area;
FIG. 10 is a diagram illustrating the operation of stacking multiple knowledge in the knowledge base.
Detailed Description
The technical scheme of the invention is explained in detail in the following with the accompanying drawings:
referring to fig. 1, a data-driven low-voltage distribution substation electricity stealing analysis method includes the following steps:
s1: zone Z to be analyzed * Performing data filling processing to calculate the region Z to be analyzed * The data characteristic of (1).
Specifically, the hourly power data of all the electric meters in the area to be analyzed are recorded as Z * The data period is a minimum of 7 days.
Preferably, the zone Z to be analyzed * The method of performing data stuffing processing is as follows:
s11: zone to be analyzed Z * The data matrix form of (a) is:
wherein (X) n ) m Representing the electric quantity of the electric meter, n is the electric meter identification under the transformer area, m is the moment of electric quantity data, and X 0 In particular to a table area general table;
longitudinal axis direction (X) 0 ) i (X 1 ) i …(X n ) i If the number of the missing data exceeds the set value (90%), deleting the electric quantity data at the moment, and turning to the step S12, otherwise, directly executing the step S12.
S12: for missing data in the matrix, the missing data is in the horizontal axis direction (X) corresponding to the missing data i ) 1 (X i ) 2 …(X i ) m The missing value is filled by polynomial interpolation, and the specific method is as follows:
the moment and the electric quantity value of the electric quantity data are regarded as point coordinates on a two-dimensional plane, the moment of the electric quantity data is an abscissa, the electric quantity value is an ordinate, and the electric quantity data of k moments are expressed as (x) 1 ,y 1 )(x 1 ,y 1 )…(x k ,y k ) From the mathematical theorem, it can be seen that there must be a k-1 degree polynomial such that (x) 1 ,y 1 )(x 1 ,y 1 )…(x k ,y k ) This holds true after the polynomial is introduced. Constructing a k-1 degree polynomial, and enabling known k coordinates to be established after being substituted into the k-1 degree polynomial, wherein the k-1 degree polynomial is as follows:
wherein L (x) is an interpolation expression, L j (x) The method is characterized in that the method is an interpolation basis function, x is a value of data time, y is a value of electric quantity, j represents the number of summations, and if a k-1 th-order polynomial is provided with k-1 summations; i represents the serial number of the coordinate, k cannot represent a linear rule if the value of k is too large, and k =24 is taken in combination with the service condition.
In [1, k ]]Introduces the (k + 1) th coordinate (x) in the range domain k+1 ,y k+1 ) Using said k-1 th order polynomial to bring in the coordinates (x) k+1 ,y k+1 ) Solving the corresponding electric quantity value to complete the first-order polynomial interpolation filling;
s13: filling all missing data according to the method in the step S12 in sequence to finish the platform zone Z to be analyzed * The data of (2) is updated.
After the updating is finished, extracting the zone Z to be analyzed * The table code, voltage/current, power, SOE event data of the middle and hour electric quantity data in the same period of time, and the data characteristics of the users under the station area are calculated, as shown in fig. 2, including: voltage (loss of voltage, loss of phase and out of limit), current (loss of current and negative current), power (excessive fluctuation and continuous reduction), table code (sudden increase and sudden decrease), event (power failure and uncovering), and the like. As shown in fig. 3, the station zone Z is obtained by calculation * The following user data characteristics.
S2: statistics of the area Z to be analyzed * Similar station zone Z of i And constructing a line loss identification model and decomposing line loss characteristics of the transformer area to be analyzed.
Preferably, the statistical analysis platform zone Z * Similar station zone Z of i The method comprises the following steps:
extracting a plurality of normal line loss distribution areas in batch from a service system, and taking the hour electric quantity data of all electric meters under the distribution areas as similar distribution area data, similar distribution areas andthe regions to be analyzed have certain similarity, and mainly comprise: geographical location, load factor, residential composition, commercial electricity proportion, etc., as shown in fig. 4, and then calculating each distribution area and the distribution area to be analyzed Z using cosine similarity * The distance between:
wherein alpha represents a station zone Z * Is the zone Z, e.g. α = (0.2, 1,0, 4) * Beta represents a batch-extracted platform area vector in the service system, alpha i 、β i Respectively, the components in the vectors alpha, beta; and sequentially carrying out similarity calculation on alpha and a station area vector beta in a service system, selecting the station areas represented by L (L = 5) vectors nearest to the alpha as similar station areas, and marking as Z i . The similarity range is [ -1,1]1 means the same direction, -1 means the opposite direction, and 0 means that the two vectors are independent.
Further, according to the step S1, the zone Z to be analyzed * Method for filling data into similar region Z i And performing data filling processing.
The method for constructing the line loss identification model comprises the following steps:
s21: as shown in fig. 5, line loss characteristics of all electricity stealing behavior types (short circuit type, modified parameter, modified structure type, electricity utilization without meter type, fault type) are abstracted into the following three types:
i: the electricity stealing users have a correlation with the line loss electricity, when the related operations such as the parameters or the structure of the ammeter are modified, the metering of the ammeter is slowed down or the precision is reduced, and at the moment, the line loss electricity of the transformer area has an obvious correlation with the electricity stealing users;
j: the electricity stealing users have no correlation between electricity quantity and line loss electricity quantity, when the electricity stealing users perform correlation operations such as flying line electricity utilization in front of the meter, the line loss electricity quantity is completely generated by an independent electricity utilization equipment set, the line loss electricity quantity only has electricity utilization characteristics, and the line loss electricity quantity has no correlation with the electricity consumption of the users;
k: the line loss approaches a constant value.
The unimorph line loss sequence expression is as follows:
I＝{a 1 (X 1 ) 1 +…+a n (X n ) 1 ,...,a 1 (X 1 ) m +…+a n (X n ) m }
J＝{b 1 ,b 2 ,...,b m }
wherein (X) n ) m For the zone Z to be analyzed * The electric quantity value of (a) is an addition sequence of the electric quantity of the electric meter multiplied by the coefficient, J is an independent power consumption independent sequence, K is an electric quantity constant sequence, considering the intermittent power consumption, a few 0 s can be allowed to exist, and n (X n ) m as elements in the line loss sequence I, the line loss values are represented, b m The line loss value of the line loss sequence J is shown, and {0, c } the line loss value of the line loss sequence K is shown.
S22: the method comprises the following steps of constructing a transformer area line loss model, dividing a transformer area line loss value into three components of I, J and K, and decomposing the characteristics of transformer area line loss, wherein the characteristics are as follows:
a j ≥0,b j ≥0,c≥0
wherein eta is 1 、η 2 、η 3 Is a hyper-parameter and is used for restricting the space of a search solution, m is the moment of electric quantity data, D represents a station area Z to be analyzed * The line loss electric quantity is obtained by subtracting all user electric meters in the distribution area from the total electric quantity, and is recorded as
d LIP To move the multiline position distance.
J = { b = { [ b ] 1 ,b 2 ,...,b m Split into sub-subsequences J sub A mixture of J and sub similar distribution area Z i The moving multiline position distance is calculated by the curve sequence of the electric quantity in (1), d LIP The calculation formula is as follows:
wherein, as shown in FIG. 6 p Is J sub And region Z i Intersection of medium electric quantity curve sequences, area p Is J sub And platform zone Z i Area of polygonal Area surrounded by medium electric quantity curve when Area p When both are 0, it indicates that the two tracks coincide without a gap.
The specific calculation method comprises the following steps: considering only the trend characteristics of the sequences, regardless of the value of the sequences, in turn by Z i Sliding a certain J in the vertical and horizontal directions by taking the point corresponding to each electric quantity value on the medium electric quantity curve sequence as a reference sub Sequence, and calculating J sub And Z i Recording the minimum distance matched in the sliding process; all J are put together sub And summing the matched minimum distances to serve as a constraint term of the line loss model.
The method for decomposing the line loss characteristics of the transformer area to be analyzed comprises the following steps: minLoss (a, b and c) is a non-explicit expression optimization problem, a meta-heuristic algorithm is used for solution space search, the accuracy of solution is not required, the specific value of each line loss characteristic does not need to be calculated, only qualitative results are needed, the numerical value of each line loss characteristic of I, J and K is judged when the algorithm converges, and if a is a j Tending towards 0 for a characteristic without I, b j Trend 0 represents the characteristic without J, trend c is 0 represents the characteristic without K, and the reverse represents the existence of the corresponding line loss characteristic,thereby obtaining the line loss characteristics of the transformer area.
The user in the transformer area inherits the line loss characteristics of the line loss electric quantity of the transformer area, and the transformer area Z is obtained * The subscriber line loss characteristics below, as shown in fig. 7.
S3: and constructing a knowledge base and defining logic derivation rules of the knowledge base.
As shown in fig. 8, the method for constructing the knowledge base includes: analyzing and collecting data under the existing system, comprising the following steps: and counting the data characteristics of the known electricity stealing users and the suspected electricity stealing users under the system, performing knowledge creation by combining power service knowledge, and establishing the relationship from the data characteristics and the line loss characteristics to the electricity stealing event to form a knowledge base. Considering the electricity utilization situation when an electricity stealing event occurs, the correlation line loss characteristics in the knowledge base are carried out according to the following principle:
A. when the line loss electricity quantity and the user electricity consumption quantity are generated by the same group of electric equipment in the same period, the line loss characteristic is I;
B. when the line loss electricity is generated by completely independent electric equipment, the line loss characteristic is J;
C. when the line loss electric quantity can be manually controlled or intervened, the line loss characteristic is K;
and if one electricity stealing event can meet a plurality of the A, the B and the C, forming a plurality of corresponding knowledge of the electricity stealing event.
In the knowledge base, 0 represents the statistics without the characteristic, 1 represents the statistics times of 1-3 times, 2 represents the statistics times of more than 3 times, and I, J and K are data values of the line loss characteristic. The knowledge base is a multi-classification decision set in nature and has uncertainty, namely, multiple pieces of knowledge can exist in the same electricity stealing event, which accords with the characteristics of electricity stealing behaviors, and the same electricity stealing behaviors can show different characteristics.
The logic derivation rule of the knowledge base comprises the following three results when a user steals the electricity sample is judged:
firstly, the electricity stealing sample certainly belongs to a knowledge base set, namely, the line loss characteristic of the electricity stealing sample can be formed by combining one or more pieces of knowledge in the knowledge base, and the user is judged to be a suspected electricity stealing user;
secondly, the electricity stealing samples may or may not belong to a knowledge base set, description is carried out through approximate knowledge, the matching degree is calculated according to the column in the knowledge base as an attribute, a plurality of knowledge combinations with the matching degree larger than a preset value are selected, and the possibility that the user is a suspected electricity stealing user is judged according to the electricity stealing event with the domination position in the knowledge combinations;
thirdly, the electricity stealing samples do not necessarily belong to the knowledge base set, and the user eliminates the possibility of electricity stealing.
Further, the logic derivation rule of the knowledge base further comprises an overlap operator defining the line loss characteristics of I, J and K:
where, n represents the simultaneous occurrence of two data sequences, the above formula describes the relationship when I, J, K are combined:
when I and J are superposed (namely, irrelevant relation superposition correlation relationship), a new uncorrelated relation J between electricity stealing users and line loss electricity is generated * ；
When I and K are superposed, the Y-axis component is increased, the correlation is not changed, but the coefficient of the correlation is increased, so that a new correlation I between the electricity stealing user electricity quantity and the line loss electricity quantity is generated * ；
When J is superposed with K, J can only pass through the trendAnd (4) performing feature analysis, increasing Y-axis components for the trend, not changing the trend features, and keeping the superposition result to be J.
When a user electricity stealing sample is judged, if the line loss characteristics of the electricity stealing sample can be formed by combining a plurality of pieces of knowledge in a knowledge base, during superposition operation of the plurality of pieces of knowledge, the line loss characteristics are subjected to superposition operation according to the definition of a superposition operation operator, data characteristics are subjected to operation according to a large value domination principle, corresponding electricity stealing events are superposed to form an array set, and the user is judged to be a suspected electricity stealing user with various electricity stealing behaviors.
S4: to-be-analyzed station region Z * And combining the line loss characteristics and the data characteristics of the lower user to form a sample vector group, and performing electricity stealing analysis by using the knowledge base.
With a zone Z to be analyzed * For example, as shown in fig. 9, the line loss identification may determine that there are a plurality of decomposed line loss features, and the associated line loss features are combined with the user data features to form a sample vector set. The data characteristics of the user A are combined with the I and K line loss characteristics of the line loss identification decomposition to form two sample vectors.
When performing line loss analysis, taking the user a as an example, the following details are provided:
(1) The sample vector of the user A can be completely matched by one piece of knowledge in the knowledge base, and the matched user A is a suspected electricity stealing user;
the sample vector of the user A can not be obtained by matching one knowledge in the knowledge base, but can be represented by a plurality of pieces of knowledge n, and the two line loss characteristics at the time of the knowledge n are subjected to superposition operation, such as
The principle that the data characteristics are dominated by larger values, such as 0 ≧ 2= max (0, 2), the electricity stealing events are superimposed to form an array set, as shown in fig. 10. At this time, the user a is a suspected electricity stealing user with 2 electricity stealing behaviors, which is in accordance with the actual situation, and the user a can be matched as the suspected electricity stealing user.
(2) The sample vector of the user A cannot be represented by one or more pieces of knowledge N in the knowledge base, at the moment, a plurality of groups of approximations are represented by the plurality of pieces of knowledge N, then the matching degree is calculated by taking columns as attributes, a plurality of knowledge combinations with the matching degree of more than 70% (matching degree = number of matched columns/total number of matched columns) are selected, the electricity stealing event sets of the plurality of combinations meeting the matching degree are observed, dominant electricity stealing events (the occurrence times are far larger than other events) in the plurality of combinations are selected, and the matched user A is a suspected electricity stealing user.
(3) And if the user A cannot meet the matching mechanism, judging that the user A is a normal electricity utilization user.
The method provides a method system for observing and analyzing the electricity stealing capacity of the transformer area from the data perspective, and the analysis capability of the electricity stealing capacity can be continuously optimized by iteratively updating the knowledge base, so that the method is suitable for the future environment.
Claims (9)
1. A data-driven low-voltage distribution station area electricity stealing analysis method is characterized by comprising the following steps:
s1: calculating the area Z to be analyzed * The data characteristic of (a);
s2: statistics of the region Z to be analyzed * Similar station zone Z of i Constructing a line loss identification model, and decomposing line loss characteristics of the distribution room to be analyzed;
s3: constructing a knowledge base, and defining logic derivation rules of the knowledge base;
s4: to-be-analyzed station region Z * And combining the line loss characteristics and the data characteristics of the lower user to form a sample vector group, and performing electricity stealing analysis by using the knowledge base.
2. The data-driven-based low-voltage distribution substation electricity stealing analysis method according to claim 1, wherein the method for constructing the line loss identification model in step S2 comprises the following steps:
s21: line loss characteristics of all electricity stealing behavior types are abstracted into the following three types:
i: the electricity stealing users have a correlation between electricity and line loss electricity;
j: the electricity stealing users have no correlation between electricity and line loss electricity;
k: the line loss electric quantity is close to a constant value;
the line loss sequence expression is as follows:
I＝{a 1 (X 1 ) 1 +…+a n (X n ) 1 ,...,a 1 (X 1 ) m +…+a n (X n ) m }
J＝{b 1 ,b 2 ,...,b m }
wherein (X) n ) m For the zone Z to be analyzed * The electric quantity value of the electric meter is I, the electric quantity of the electric meter is multiplied by the addition sequence of the coefficients, J is an independent power consumption independent sequence, and K is an electric quantity constant sequence;
s22: constructing a platform area line loss model:
a j ≥0,b j ≥0,c≥0
wherein eta is 1 、η 2 、η 3 Is a hyper-parameter and is used for restricting the space of a search solution, m is the moment of electric quantity data, D represents a station area Z to be analyzed * Line loss of electricity, d LIP Moving the multiline position distance;
j = { b = { [ b ] 1 ,b 2 ,...,b m Split into sub-subsequences J sub A mixture of J and sub similar station zone Z i The moving multiline position distance is calculated by the curve sequence of the electric quantity in (1), d LIP The calculation formula is as follows:
wherein, t p Is J sub And platform zone Z i Intersection of medium electric quantity curve sequences, area p Is J sub And region Z i The area of a polygonal area surrounded by the medium electric quantity curve;
in sequence with Z i The point corresponding to each electric quantity value on the medium electric quantity curve sequence is taken as a reference, and a certain J slides in the vertical and horizontal directions sub Sequence, and calculating J sub And Z i Recording the minimum distance matched in the sliding process;
all J are put together sub And summing the matched minimum distances to serve as a constraint term of the line loss model.
3. The data-driven-based low-voltage distribution substation area electricity stealing analysis method according to claim 2, characterized in that: s2, the method for decomposing the line loss characteristics of the transformer area to be analyzed comprises the following steps: and (3) performing solution space search on minLoss (a, b and c) by using a meta-heuristic algorithm, and judging the numerical values of the line loss characteristics of I, J and K when the algorithm converges to obtain the line loss characteristics of the transformer area.
4. The data-driven-based low-voltage distribution substation electricity stealing analysis method according to claim 2, wherein the method for constructing the knowledge base in step S3 comprises the following steps: the method comprises the following steps of counting the data characteristics of known electricity stealing users and suspected electricity stealing users under a system, creating knowledge by combining power service knowledge, establishing the relation from the data characteristics and line loss characteristics to electricity stealing events, and forming a knowledge base, wherein the association of the line loss characteristics in the knowledge base is carried out according to the following principle:
A. when the line loss electricity quantity and the user electricity consumption quantity are generated by the same group of electric equipment in the same period, the line loss characteristic is I;
B. when the line loss electricity quantity is generated by completely independent electric equipment, the line loss is characterized by J;
C. when the line loss electric quantity can be manually controlled or intervened, the line loss characteristic is K;
if one electricity stealing event can meet a plurality of the A, B and C, forming a plurality of corresponding knowledge of the electricity stealing event;
the logic derivation rule of the knowledge base comprises the following three results when a user steals the electricity sample is judged:
firstly, the electricity stealing sample certainly belongs to a knowledge base set, namely, the line loss characteristic of the electricity stealing sample can be formed by combining one or more pieces of knowledge in the knowledge base, and the user is judged to be a suspected electricity stealing user;
secondly, the electricity stealing samples may or may not belong to a knowledge base set, description is carried out through approximate knowledge, the matching degree is calculated according to the column in the knowledge base as an attribute, a plurality of knowledge combinations with the matching degree larger than a preset value are selected, and the possibility that the user is a suspected electricity stealing user is judged according to the electricity stealing event with the domination position in the knowledge combinations;
thirdly, the electricity stealing samples do not necessarily belong to the knowledge base set, and the user eliminates the possibility of electricity stealing.
5. The data-driven-based low-voltage distribution substation area electricity stealing analysis method according to claim 4, wherein: and S3, the logic derivation rule of the knowledge base further comprises an overlap operator for defining line loss characteristics of I, J and K:
where n denotes that two data sequences occur simultaneously,
when I and J are superposed, a new uncorrelated relation J between electricity stealing users and line loss electricity is generated * ，
When the I and the K are superposed, a new relation I between the electricity stealing user electricity quantity and the line loss electricity quantity is generated * ，
The trend characteristic is not changed when J and K are superposed, and the superposition result is still J;
when a user electricity stealing sample is judged, if the line loss characteristics of the electricity stealing sample can be formed by combining a plurality of pieces of knowledge in a knowledge base, during superposition operation of the plurality of pieces of knowledge, the line loss characteristics are subjected to superposition operation according to the definition of a superposition operation operator, data characteristics are subjected to operation according to a large value domination principle, corresponding electricity stealing events are superposed to form an array set, and the user is judged to be a suspected electricity stealing user with various electricity stealing behaviors.
6. The data-driven-based low-voltage distribution substation area electricity stealing analysis method according to claim 1, wherein: the step S1 is to calculate the area Z to be analyzed * Before the data characteristic of (2) the zone to be analyzed Z is first treated * And performing data filling processing by the following method:
s11: zone to be analyzed Z * The data matrix form of (a) is:
wherein (X) n ) m Representing the electric quantity of the electric meter, n is the electric meter identification under the transformer area, m is the moment of electric quantity data, and X 0 In particular to a table area general table;
longitudinal axis direction (X) 0 ) i (X 1 ) i …(X n ) i In the above-mentioned manner,if the number of the missing data exceeds the set value, deleting the electric quantity data at the moment, and turning to the step S12, otherwise, directly executing the step S12;
s12: for missing data in the matrix, the missing data is in the horizontal axis direction (X) corresponding to the missing data i ) 1 (X i ) 2 …(X i ) m The missing value is filled by polynomial interpolation, and the specific method is as follows:
the moment and the electric quantity value of the electric quantity data are regarded as point coordinates on a two-dimensional plane, the moment of the electric quantity data is an abscissa, the electric quantity value is an ordinate, and the electric quantity data of k moments are expressed as (x) 1 ,y 1 )(x 1 ,y 1 )...(x k ,y k ) And constructing a k-1 degree polynomial so that known k coordinates are substituted into the k-1 degree polynomial, wherein the k-1 degree polynomial is as follows:
wherein L (x) is an interpolation expression, L j (x) The sum is an interpolation basis function, x is a value of data time, y is a value of electric quantity, and j represents the sum number;
in [1, k ]]Introduces the (k + 1) th coordinate (x) in the range domain k+1 ,y k+1 ) Using said k-1 th order polynomial to bring in the coordinates (x) k+1 ,y k+1 ) Solving the corresponding electric quantity value to complete the first-order polynomial interpolation filling;
s13: filling all missing data according to the method in the step S12 in sequence to finish the distribution room Z to be analyzed * The data of (3) is updated.
7. The data-driven-based low-voltage distribution substation electricity stealing analysis method according to claim 6, wherein: step S1 is to-be-analyzed transformer area Z * When the data filling-in process is performed,k＝24。
8. the data-driven-based low-voltage distribution substation area electricity stealing analysis method according to claim 1, wherein: s2, counting the distribution area Z to be analyzed * Similar station zone Z of i The method comprises the following steps: extracting platform region vectors in batches from a service system, and calculating Z of each platform region and a platform region to be analyzed by using cosine similarity * The distance between:
wherein alpha represents a plateau Z * Beta represents a batch-extracted platform area vector in the service system, alpha i 、β i Each component in the vectors α, β, respectively;
sequentially carrying out similarity calculation on alpha and a distribution room vector beta in a service system, and selecting a distribution room Z to be analyzed * Taking L nearest distribution areas as similar distribution areas Z i 。
9. The data-driven-based low-voltage distribution substation electricity stealing analysis method according to claim 6, wherein: the step S2 further comprises the step of analyzing the zone Z to be analyzed according to the step S1 * Method for filling data into similar distribution area Z i And performing data filling processing.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210976257.5A CN115330202B (en) | 2022-08-15 | 2022-08-15 | Low-voltage distribution area electricity larceny analysis method based on data driving |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210976257.5A CN115330202B (en) | 2022-08-15 | 2022-08-15 | Low-voltage distribution area electricity larceny analysis method based on data driving |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115330202A true CN115330202A (en) | 2022-11-11 |
| CN115330202B CN115330202B (en) | 2023-07-11 |
Family
ID=83922984
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210976257.5A Active CN115330202B (en) | 2022-08-15 | 2022-08-15 | Low-voltage distribution area electricity larceny analysis method based on data driving |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115330202B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2014090018A1 (en) * | 2012-12-12 | 2014-06-19 | 天津市电力公司 | Iec61850 standard-based intelligent power distribution terminal |
| CN105373877A (en) * | 2015-09-14 | 2016-03-02 | 江苏南瑞通驰自动化系统有限公司 | Electricity utilization trend anomaly suspicion analysis and anti-electric-larceny monitoring system |
| US20170330086A1 (en) * | 2016-05-12 | 2017-11-16 | Electronics And Telecommunications Research Institute | Method and system for analyzing data |
| CN111930802A (en) * | 2020-08-01 | 2020-11-13 | 青岛鼎信通讯股份有限公司 | Anti-electricity-stealing analysis method based on Lasso analysis |
| CN114019205A (en) * | 2021-07-16 | 2022-02-08 | 国家电网有限公司技术学院分公司 | Electricity stealing identification method and system |
| CN114862139A (en) * | 2022-04-19 | 2022-08-05 | 国网江苏省电力有限公司南通供电分公司 | Data-driven transformer area line loss rate abnormity diagnosis method |
-
2022
- 2022-08-15 CN CN202210976257.5A patent/CN115330202B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2014090018A1 (en) * | 2012-12-12 | 2014-06-19 | 天津市电力公司 | Iec61850 standard-based intelligent power distribution terminal |
| CN105373877A (en) * | 2015-09-14 | 2016-03-02 | 江苏南瑞通驰自动化系统有限公司 | Electricity utilization trend anomaly suspicion analysis and anti-electric-larceny monitoring system |
| US20170330086A1 (en) * | 2016-05-12 | 2017-11-16 | Electronics And Telecommunications Research Institute | Method and system for analyzing data |
| CN111930802A (en) * | 2020-08-01 | 2020-11-13 | 青岛鼎信通讯股份有限公司 | Anti-electricity-stealing analysis method based on Lasso analysis |
| CN114019205A (en) * | 2021-07-16 | 2022-02-08 | 国家电网有限公司技术学院分公司 | Electricity stealing identification method and system |
| CN114862139A (en) * | 2022-04-19 | 2022-08-05 | 国网江苏省电力有限公司南通供电分公司 | Data-driven transformer area line loss rate abnormity diagnosis method |
Non-Patent Citations (2)
| Title |
|---|
| 唐冬来;刘友波;熊智临;马铁丰;苏童;: "基于时空关联矩阵的配电台区反窃电预警方法", 电力系统自动化, no. 19, pages 247 - 260 * |
| 金晟;苏盛;曹一家;薛阳;杨艺宁;刘厦: "基于格兰杰归因分析的高损台区窃电检测", 电力系统自动化, vol. 44, no. 023, pages 82 - 89 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115330202B (en) | 2023-07-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111199016B (en) | Daily load curve clustering method for improving K-means based on DTW | |
| Wang et al. | Population-based intelligent search in reliability evaluation of generation systems with wind power penetration | |
| Yu et al. | Mathematical analysis and solutions for multi-objective line-cell conversion problem | |
| Cheng et al. | Air traffic control using genetic search techniques | |
| CN111612650B (en) | DTW distance-based power consumer grouping method and system | |
| CN108985380B (en) | Point switch fault identification method based on cluster integration | |
| CN114519514B (en) | Low-voltage transformer area reasonable line loss value measuring and calculating method, system and computer equipment | |
| CN112381673B (en) | Park electricity utilization information analysis method and device based on digital twin | |
| CN112308298B (en) | Multi-scenario performance index prediction method and system for semiconductor production line | |
| CN110991527A (en) | Similarity threshold determination method considering voltage curve average fluctuation rate | |
| CN110610121A (en) | Small-scale source load power abnormal data identification and restoration method based on curve clustering | |
| CN110348540B (en) | Clustering-based method and device for screening transient power angle stability faults of power system | |
| CN113361801A (en) | Hybrid rolling short-term load prediction method based on GAN network and semi-supervised regression | |
| CN112684295A (en) | Power distribution network fault line selection method and system under high permeability based on similarity separation degree | |
| CN108808657B (en) | Short-term prediction method for power load | |
| CN111370055B (en) | Intron retention prediction model establishment method and prediction method thereof | |
| CN109784362A (en) | A kind of DGA shortage of data value interpolating method based on iteration KNN and interpolation priority | |
| CN113177357A (en) | Transient stability assessment method for power system | |
| CN115330202A (en) | Data-driven low-voltage distribution station area electricity stealing analysis method | |
| CN117609818A (en) | Power grid association relation discovery method based on clustering and information entropy | |
| CN115296295B (en) | Low-voltage distribution transformer area topology identification method and system based on cluster analysis and graph theory | |
| CN113609109B (en) | Automatic scene information generation method based on data twinning | |
| CN112363012A (en) | Power grid fault early warning device and method | |
| CN109711913A (en) | A kind of electric car supply and demand difference prediction technique and charging pile method of allocation plan | |
| CN116068303A (en) | Private capacity-increasing on-line monitoring method for special transformer based on data driving |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |