CN116718979B - Smart electric meter operation error measurement method and system - Google Patents
Smart electric meter operation error measurement method and system Download PDFInfo
- Publication number
- CN116718979B CN116718979B CN202310987000.4A CN202310987000A CN116718979B CN 116718979 B CN116718979 B CN 116718979B CN 202310987000 A CN202310987000 A CN 202310987000A CN 116718979 B CN116718979 B CN 116718979B
- Authority
- CN
- China
- Prior art keywords
- meter
- user
- ammeter
- representing
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000691 measurement method Methods 0.000 title claims description 11
- 238000005259 measurement Methods 0.000 claims abstract description 70
- 238000005457 optimization Methods 0.000 claims abstract description 37
- 238000013528 artificial neural network Methods 0.000 claims abstract description 35
- 238000000034 method Methods 0.000 claims abstract description 28
- 230000005540 biological transmission Effects 0.000 claims abstract description 13
- 230000002159 abnormal effect Effects 0.000 claims abstract description 6
- 238000007781 pre-processing Methods 0.000 claims abstract description 5
- 238000004134 energy conservation Methods 0.000 claims abstract description 4
- 238000012546 transfer Methods 0.000 claims description 16
- 239000013598 vector Substances 0.000 claims description 16
- 230000008569 process Effects 0.000 claims description 11
- 230000003190 augmentative effect Effects 0.000 claims description 7
- 230000000737 periodic effect Effects 0.000 claims description 7
- 239000011159 matrix material Substances 0.000 claims description 5
- 230000008859 change Effects 0.000 claims description 4
- 101150057833 THEG gene Proteins 0.000 claims description 2
- 230000004931 aggregating effect Effects 0.000 claims description 2
- 230000005611 electricity Effects 0.000 claims description 2
- 230000006870 function Effects 0.000 description 19
- 238000004364 calculation method Methods 0.000 description 11
- 238000012545 processing Methods 0.000 description 6
- 230000000694 effects Effects 0.000 description 4
- 238000000547 structure data Methods 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000009413 insulation Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 238000002372 labelling Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 238000003062 neural network model Methods 0.000 description 2
- 230000005856 abnormality Effects 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 230000003416 augmentation Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000013450 outlier detection Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R35/00—Testing or calibrating of apparatus covered by the other groups of this subclass
- G01R35/04—Testing or calibrating of apparatus covered by the other groups of this subclass of instruments for measuring time integral of power or current
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Remote Monitoring And Control Of Power-Distribution Networks (AREA)
Abstract
The invention discloses a method and a system for measuring operation errors of an intelligent ammeter, which belong to the technical field of measuring electric variables, wherein the method comprises the following steps: acquiring a smart meter topological structure of a station area through a graph neural network, and transmitting the smart meter topological structure through a message transmission neural network; according to the topological structure of the intelligent electric meter, constructing an intelligent electric meter operation error measurement model based on an energy conservation law; acquiring intelligent ammeter data of a plurality of periods, and preprocessing the intelligent ammeter data to remove abnormal data; establishing a smart meter operation error measurement equation set according to the preprocessed smart meter data; solving an intelligent ammeter operation error measurement equation set based on a sparse optimization mode to obtain the relative error and the line loss rate of each user ammeter; when the relative error of the user ammeter is larger than the preset error value or the line loss rate is in the preset range, determining that the user ammeter has faults, and giving an alarm.
Description
Technical Field
The invention belongs to the technical field of measuring electric variables, and particularly relates to a method and a system for measuring operation errors of an intelligent ammeter.
Background
With the continuous development of social economy, the scale of the smart grid is gradually enlarged, and more than 5 hundred million electric meters are running in China according to statistics. The intelligent ammeter is one of basic equipment for data acquisition of an intelligent power grid (particularly an intelligent power distribution network), and takes on the tasks of original electric energy data acquisition, metering and transmission, so that the judgment of the operation error state of the intelligent ammeter is very critical.
Currently, the calculation of the operation error of the smart meter is mainly performed by solving a linear equation set based on the principle of conservation of electric energy, however, as the number of the smart meters continues to increase, the topology of the smart meter changes frequently, each time the topology of the smart meter changes, a tester is required to reset parameters of the equation set, a new linear equation set is established, and as the equation in the linear equation set increases, the solving speed also becomes slow, the error measurement efficiency is low, and meanwhile, the probability of calculating an invalid solution increases, so that the accuracy of error measurement is reduced.
Disclosure of Invention
In order to solve the technical problems that in the prior art, when the running error of the intelligent ammeter is calculated by solving a linear equation set based on an electric energy conservation principle, a tester is required to reset equation set parameters each time the topology structure of the intelligent ammeter changes, a new linear equation set is established, the linear equation set is sick due to the increase of equations in the linear equation set, the solving speed is slow, the error measurement efficiency is low, meanwhile, the probability of calculating an invalid solution is increased, and the accuracy of error measurement is reduced.
First aspect
The invention provides a method for measuring the operation error of an intelligent ammeter, which comprises the following steps:
s101: acquiring a smart meter topological structure of a platform region through a graph neural network, and transmitting the smart meter topological structure through a message transmission neural network, wherein the smart meter topological structure comprises a plurality of summary tables and a plurality of user meters connected with the summary tables;
s102: according to the topological structure of the intelligent electric meter, based on the law of conservation of energy, an intelligent electric meter operation error measurement model is constructed:
wherein y (i) represents the measured value of the i-th periodic total table, x j (i) A measurement value representing the jth user meter of the ith period, n representing the number of user meters, ε j Representing the relative error of the jth user ammeter, E (i) representing the energy loss of the ith period, the energy loss being calculated by the line loss rate;
s103: acquiring intelligent ammeter data of a plurality of periods, and preprocessing the intelligent ammeter data to remove abnormal data;
s104: establishing a smart meter operation error measurement equation set according to the preprocessed smart meter data;
s105: solving an intelligent ammeter operation error measurement equation set based on a sparse optimization mode to obtain the relative error and the line loss rate of each user ammeter;
s106: when the relative error of the user ammeter is larger than the preset error value or the line loss rate is in the preset range, determining that the user ammeter has faults, and giving an alarm.
Second aspect
The invention provides a smart meter operation error measurement system for executing the smart meter operation error measurement method in the first aspect.
Compared with the prior art, the invention has at least the following beneficial technical effects:
(1) According to the invention, even if the number of the intelligent electric meters is continuously increased, the topology structure of the intelligent electric meters is changed, the topology structure of the intelligent electric meters in the platform area can be obtained in real time through the graph neural network, and the topology structure of the intelligent electric meters is transmitted through the message transmission neural network, so that the operation error measurement model of the intelligent electric meters is automatically updated, a tester is not required to reset equation set parameters and newly establish a new linear equation set, and the error measurement efficiency is improved.
(2) According to the invention, based on a sparse optimization mode, the solution of the linear equation set can be thinned by introducing prior information, namely, a plurality of variables in the solution are zero, and the original large-scale linear equation set problem can be converted into a smaller-scale optimization problem, so that only a few variables need to be calculated in the process of solving the intelligent ammeter operation error measurement equation set, the calculation complexity and the difficulty of solving are greatly reduced, and the error measurement efficiency is improved.
(3) In the invention, based on a sparse optimization mode, the algorithm can be promoted to calculate a sparser solution through regularization constraint, so that the calculation of an invalid solution is avoided, and the accuracy of error measurement is improved.
Drawings
The above features, technical features, advantages and implementation of the present invention will be further described in the following description of preferred embodiments with reference to the accompanying drawings in a clear and easily understood manner.
FIG. 1 is a schematic flow chart of a method for measuring the operation error of a smart meter;
fig. 2 is a schematic structural diagram of a method for measuring operation errors of a smart meter.
Detailed Description
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following description will explain the specific embodiments of the present invention with reference to the accompanying drawings. It is evident that the drawings in the following description are only examples of the invention, from which other drawings and other embodiments can be obtained by a person skilled in the art without inventive effort.
For simplicity of the drawing, only the parts relevant to the invention are schematically shown in each drawing, and they do not represent the actual structure thereof as a product. Additionally, in order to simplify the drawing for ease of understanding, components having the same structure or function in some of the drawings are shown schematically with only one of them, or only one of them is labeled. Herein, "a" means not only "only this one" but also "more than one" case.
It should be further understood that the term "and/or" as used in the present specification and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes such combinations.
In this context, it should be noted that the terms "mounted," "connected," and "connected" are to be construed broadly, and may be, for example, fixedly connected, detachably connected, or integrally connected, unless otherwise explicitly stated and defined. Either mechanically or electrically. Can be directly connected or indirectly connected through an intermediate medium, and can be communication between two elements. The specific meaning of the above terms in the present invention will be understood in specific cases by those of ordinary skill in the art.
In addition, in the description of the present invention, the terms "first," "second," and the like are used merely to distinguish between descriptions and are not to be construed as indicating or implying relative importance.
Example 1
In one embodiment, referring to fig. 1 of the specification, a flow chart of a method for measuring operation errors of a smart meter provided by the invention is shown. Referring to fig. 2 of the specification, a schematic structural diagram of a method for measuring operation errors of a smart meter is shown.
The invention provides a method for measuring the operation error of an intelligent ammeter, which comprises the following steps:
s101: the intelligent ammeter topological structure of the area is obtained through the graphic neural network, and is transmitted through the message transmission neural network, and comprises a plurality of summary tables and a plurality of user ammeter connected with the summary tables.
Among these, the graph neural network (Graph Neural Network, GNN) is a type of machine learning model for processing graph structure data. Unlike conventional neural networks, which primarily process vector or matrix data, graph neural networks are dedicated to processing graph data that includes nodes and edges.
In a possible implementation manner, the method for obtaining the intelligent ammeter topology structure of the area through the graph neural network specifically includes:
and obtaining the connection relation between the summary list and the user electric meter through the graph neural network.
And according to the connection relation between the summary list and the user electric meters and the connection relation between the user electric meters, constructing intelligent electric meter graph data by taking each intelligent electric meter as a node and each connection relation as an edge.
And converting nodes and edges in the intelligent ammeter graph data into low-dimensional vectors through a GCN graph embedding model to obtain embedded representation of the intelligent ammeter topological structure.
The GCN (Graph Convolutional Network, based on a graph rolling network) graph embedding model is a neural network model for processing graph structure data, and can convert node representations in the graph into low-dimensional vectors so as to realize the dimension reduction and feature extraction of the graph data.
It should be noted that, in the conventional method, the obtaining of the topology structure of the smart meter of the platform area generally needs to rely on manual labeling or manual setting, which has a large workload and is prone to error. And by adopting a graphic neural network, particularly a GCN graphic embedded model, the connection relation between the total table in the area and the user electric meter and the connection relation between the user electric meters can be automatically learned from the data. Thus, the burden of manual labeling can be reduced, and a more accurate topological structure can be obtained.
Further, even if the number of the intelligent electric meters is continuously increased, the topology structure of the intelligent electric meters is changed, the topology structure of the intelligent electric meters in the platform area can be obtained in real time through the graph neural network, the topology structure of the intelligent electric meters is obtained through the graph neural network, the embedded representation is obtained through the GCN graph embedded model, the topology structure can be automatically learned, the global feature expression is obtained, the data dimension is reduced, and the intelligent electric meter has strong generalization capability, so that a more efficient, accurate and reliable foundation is provided for analysis and application of the data of the intelligent electric meters.
Among other things, messaging neural networks (Message Passing Neural Network, MPNN) are a class of graph neural network models for processing graph structure data. MPNN is an important variant of the neural network of the graph, and the core idea is to update the characteristics and transfer the information of the nodes in the graph by means of message transfer.
In one possible implementation, the messaging neural network includes a plurality of nodes, and the smart meter topology is transferred through the messaging neural network, specifically including:
acquiring hidden states of nodes i adjacent to each node v after t times of message exchangeAnd the original hidden state of each node v +.>
Hidden states of adjacent nodes i after t times of message exchange are carried out through a message transfer functionAnd the original hidden state of each node v +.>Combining:
where m represents a message transfer function, b 1 、b 2 Indicating the amount of offset.
According to the message transfer function, obtaining the message exchanged by the node v at t+1 times, and transferring:
wherein ,representing the messages exchanged by node v at time t+1, N (v) representing the total number of nodes.
Where the message transfer function is one of the key components in the graph neural network for transferring information and features between nodes in the graph structure data. The main purpose of the message transfer function is to communicate and transfer information between one node and its neighboring nodes, thereby updating the node's feature representation.
Updating a message transfer function based on previous hidden statesAnd aggregating the messages to calculate a new hidden state +/for each node>Repeating the above process, and iterating until reaching the preset iteration times.
The intelligent ammeter topology structure is transferred through the message transfer neural network, the characteristic transfer of the nodes is realized, the neighbor node information can be adaptively learned, the multi-round iterative transfer, the parameter sharing and the efficient calculation are realized, and the global graph characteristics are obtained, so that the intelligent ammeter topology structure is more accurately, comprehensively and efficiently expressed. The characteristics enable the graph neural network to have advantages and potential when processing graph data such as the intelligent ammeter topological structure.
S102: according to the topological structure of the intelligent electric meter, based on the law of conservation of energy, an intelligent electric meter operation error measurement model is constructed:
wherein y (i) represents the measured value of the i-th periodic total table, x j (i) A measurement value representing the jth user meter of the ith period, n representing the number of user meters, ε j Representing the relative error of the jth consumer meter, E (i) represents the energy loss of the ith period, which can be calculated from the line loss rate.
It should be noted that, the error measurement model based on the law of conservation of energy is a method based on the physical principle, and can model the ammeter data from the angle of conservation of energy. The model can reflect the transmission and consumption processes of the electric energy more truly, so that the accuracy of error measurement is improved.
Furthermore, by constructing an intelligent ammeter operation error measurement model based on an energy conservation law, the authenticity and accuracy of error measurement can be improved, line loss and energy consumption are considered, the situation of a plurality of user ammeters is comprehensively considered, and a powerful basis is provided for evaluation and improvement of power consumption of a transformer area.
In one possible embodiment, the energy loss E (i) of the ith cycle is specifically:
wherein ,eM (i) Represents the self-loss of the electric energy meter in the ith period, e N (i) Represents the ith periodic leakage loss, e L (i) Indicating the i-th periodic line loss,indicating the rated power of the jth consumer meter, t (i) indicating the meter run time of the ith period, U indicating the line voltage, sigma indicating the line leakage conductivity, +.>Representing the line current of the jth user ammeter in the ith period, R j Indicating the line resistance of the jth meter to the summary, and n indicating the number of meters.
The self-loss of the electric energy meter is energy loss generated in the electric energy measuring process of the intelligent electric energy meter. The self-loss of an electric energy meter is usually caused by internal circuits, electronic components, measuring sensors, and the like. In general, the self-loss of an electric energy meter can be regarded as a fixed amount that remains unchanged.
The leakage loss refers to energy leakage caused by poor insulation or equipment failure in the process of conveying electric energy. When the equipment has a leakage problem, electric energy can be lost through a path of the insulation defect, so that energy loss is caused. Leakage losses are typically due to insulation aging, equipment damage, or ground faults.
The line loss refers to energy loss of electric energy in the power transmission line due to electric wire and cable resistance. When power is delivered from a power source to a user terminal, it passes through a series of transmission lines, which themselves have a certain resistance. Heat is generated when current passes through the resistor, resulting in energy loss, which is the line loss. Line losses typically vary with transmission distance, line material, current magnitude, and the like.
Self-loss e of electric energy meter which is kept unchanged in each period M (i) Defined as fixed loss w 0 。
Leakage loss e which will be difficult to accurately measure at each cycle N (i) And line loss e L (i) Defining as a change loss, and establishing a relationship between the change loss and a measurement value of a total table through a line loss rate:
e N (i)+e L (i)＝b 0 ·Δy(i)＝b 0 ·[y(i)-y(i-1)]；
wherein ,b0 The line loss rate is represented, and Δy (i) represents the total power consumption in the ith cycle.
The line loss rate refers to the proportion of energy loss caused by factors such as resistance in a power transmission line when electric energy is transmitted from a power supply to a user terminal.
S103: and acquiring intelligent ammeter data of a plurality of periods, and preprocessing the intelligent ammeter data to remove abnormal data.
Wherein, smart electric meter data includes: the measurement value of each summary of the continuous period, the measurement value of each user electricity meter.
Specifically, preprocessing includes data cleansing, outlier detection, missing value processing, data normalization, and the like.
In one possible implementation, S103 specifically includes substeps S1031 to S1034:
s1031: and comparing the sum of the measured values of the user electric meters in the intelligent electric meter data with the total measured value.
S1032: and (3) retaining the data of the sum of the measured values of the user electric meters smaller than or equal to the measured value of the total table, and removing the data of the sum of the measured values of the user electric meters larger than the measured value of the total table.
It should be noted that, due to the loss, the total measured value should be less than or equal to the sum of all the measured values of the user electric meters, which is in accordance with the law of conservation of energy. If the sum of the measured values of all the user electric meters in certain data is larger than the total measured value, the abnormal data is indicated, the abnormal data is removed, and the quality and the reliability of the intelligent electric meter data can be improved.
S1033: calculating the energy loss rate of each intelligent ammeter data:
wherein ,γi Energy loss rate of smart meter data representing the ith cycle, y (i) represents a measurement value of the total table of the ith cycle, x j (i) Representing the measurement value of the jth meter for the ith period, j=1, 2, …, n, n representing the number of meters.
S1034: and (3) preserving the intelligent ammeter data with the energy loss rate smaller than or equal to the preset loss rate, and removing the intelligent ammeter data with the energy loss rate larger than the preset loss rate.
The energy loss rate that is too high may be a manifestation of problems such as data anomalies and power theft. The data with the energy loss rate smaller than or equal to the preset loss rate is reserved, the data with the energy loss rate larger than the preset loss rate is removed, the data abnormality can be identified and eliminated, and the accuracy and the reliability of the data of the intelligent electric meter are ensured.
The size of the preset loss rate can be set by a person skilled in the art according to practical situations, and the invention is not limited.
Further, by calculating the energy loss rate of the intelligent ammeter data and reserving the data meeting the preset loss rate condition, more comprehensive, accurate and reliable intelligent ammeter data can be provided, and a better data base is provided for the establishment and optimization of a subsequent error measurement model.
S104: and establishing a system of intelligent ammeter operation error measurement equations according to the preprocessed intelligent ammeter data.
In one possible implementation, S104 is specifically:
the intelligent ammeter operation error measurement equation set is established through the following formula:
wherein ,aj Representing the error coefficient of the jth user meter, b 0 Represents the line loss rate, w 0 Representing the fixed loss x j (i) Representing a measured value of a jth user meter in the ith smart meter data, y (i) representing a measured value of a total table in the ith smart meter data, Δy (i) representing a total power consumption in the ith smart meter data, i=1, 2, …, m, m representing a total number of smart meter data, j=1, 2, …, n, n representing the number of user meters.
Wherein the error coefficient a j The expression of (2) is:
wherein ,εj Indicating the relative error of the jth consumer meter.
S105: and solving an intelligent ammeter operation error measurement equation set based on a sparse optimization mode to obtain the relative error and the line loss rate of each user ammeter.
The sparse optimization is an optimization method, and the aim of the sparse optimization is to enable solutions of the optimization problem to have zero elements (sparsity) as much as possible on the premise of meeting certain constraint conditions. In sparse optimization, it is desirable to find a set of optimal solutions where most elements are zero and only a few non-zero elements, which can translate the original large-scale problem into a smaller-scale optimization problem.
It should be noted that, based on the sparse optimization mode, the solution of the linear equation set can be thinned by introducing prior information, namely, many variables in the solution are zero, and the original large-scale linear equation set problem can be converted into a smaller-scale optimization problem, so that only a few variables need to be calculated in the process of solving the intelligent ammeter operation error measurement equation set, the calculation complexity and the difficulty degree of solving are greatly reduced, and the error measurement efficiency is improved.
Further, based on a sparse optimization mode, the algorithm can be prompted to calculate a sparser solution through regularization constraint, invalid solutions are prevented from being calculated, and accuracy of error measurement is improved.
In one possible embodiment, S105 specifically includes substeps S1051 to S1056:
s1051: constructing a sparse optimization model of a measurement equation set of the operation error of the intelligent ammeter:
A＝[a 1 … a j ],
W＝[w 0 … w 0 ],ΔY＝[Δy(1) … Δy(m)],Y＝[y(1) … y(m)]
wherein A represents an error coefficient vector, X represents a measured value matrix of a user ammeter, W represents a fixed loss matrix, deltaY represents a total power consumption vector, Y represents a total measured value vector, lambda represents a regularization parameter of relative error, mu represents a regularization parameter of line loss, and DeltaS is the same 1 A norm is represented by a number of norms, I.I 2 Representing the two norms, y 0 An a priori term representing line loss.
Wherein, a norm is 1 Representing the sum of the absolute values of the vector elements, two norms |. || 2 Representing the sum of squares and the open squares of the vector elements.
Regularization is a technique that introduces additional constraints or penalty terms in the optimization problem, aiming at improving the generalization ability and stability of the model. The regularization parameter λ of the relative error is introduced in order to limit the magnitude of the relative error. By adjusting the lambda, the fitting degree of the relative error and the sparsity of the model can be balanced, so that a more reasonable error coefficient is obtained. The regularization parameter mu of the lead-in line loss is used for controlling the size of the line loss rate and avoiding the line loss rate from taking an excessive or insufficient value. By adjusting the size of mu, the fitting degree and sparsity of the line loss rate can be balanced, so that the more accurate line loss rate is obtained.
S1052: fixing the line loss rate, and updating a sparse optimization model of an intelligent ammeter operation error measurement equation set:
s1053: order theg(Z)＝λ||Z|| 1 An augmented lagrangian function is constructed:
where ρ represents a penalty parameter and Q represents a penalty variable.
The augmentation Lagrangian function is to expand the original Lagrangian function, introduce additional variables and penalty parameters, and solve the optimization problem with equality and inequality constraints. In sparse optimization, an augmented lagrangian function is often used to address optimization problems with sparsity constraints.
Wherein the penalty parameter ρ plays a role in controlling the degree of constraint violation in the augmented lagrangian function. In sparse optimization, the objective of introducing an augmented lagrangian function is to translate constraints into a part of the objective function and balance the trade-off between optimization of the objective function and meeting the constraints by adjusting the penalty parameter ρ.
Wherein, penalty variable Q acts as an auxiliary constraint, and is an additional variable that penalizes the sparsity constraint.
S1054: x, Z and Q are updated according to the augmented Lagrangian function and f (X) is calculated.
Specifically, the process of alternately updating X, Z and W gradually optimizes the objective function and satisfies the constraint condition by continuous iteration.
Alternatively, the Alternate Direction Multiplier Method (ADMM) is used to update X, Z and W, and f (X) is calculated from X.
S1055: according toAn error coefficient vector a is calculated.
S1056: according to error coefficient vectors A andcalculating the relative error epsilon of the jth user ammeter j 。
In one possible implementation, S105 further comprises sub-steps S1057 to S1059:
s1057: fixing the relative error, updating a sparse optimization model of an intelligent ammeter operation error measurement equation set, solving the line loss rate, and updating the sparse optimization model to be:
s1058: order theThe second derivative h "(ΔY) of h (ΔY) is determined.
S1059: let h "(Δy) =0, calculate the line loss rate b 0 。
It should be noted that, by solving the second derivative of the objective function, information about curvature and convexity of the objective function can be provided, which helps us to optimize more effectively, find the minimum value of the objective function, and thus obtain the optimal solution of the line loss rate.
S106: when the relative error of the user ammeter is larger than the preset error value or the line loss rate is in the preset range, determining that the user ammeter has faults, and giving an alarm.
The size of the preset error value can be set by a person skilled in the art according to practical situations, and the invention is not limited.
The specific range of the preset range can be set by a person skilled in the art according to practical situations, and the invention is not limited.
It should be noted that, when the relative error of the user meter is greater than the preset error value, it often means that the measured data of the user meter is unreliable, and the user meter has a fault. On the other hand, when the line loss rate is within the preset range, it often means that there is a problem in the power line, for example, there is a situation of leakage, excessive loss, and the like, and the user ammeter has a fault.
Specifically, an alarm may be sounded by sounding an alarm, sending a short message, sending an email, etc., to notify related personnel, such as maintenance personnel, an operation and maintenance team, or a system administrator, so that they can take action in time to handle the fault.
In one possible embodiment, the method further comprises:
s107: calculating Root Mean Square Error (RMSE) of an intelligent ammeter operation error measurement model:
wherein ,εj Representing the relative error of the jth user ammeter calculated by the intelligent ammeter operation error measurement model,indicating the actual error of the jth consumer meter.
S108: when the root mean square error of the intelligent ammeter operation error measurement model is larger than an error threshold value, the regularization parameter lambda of the relative error, the regularization parameter mu of the line loss and the punishment parameter rho are adjusted, and calculation is carried out again.
The error threshold value can be set by a person skilled in the art according to practical situations, and the invention is not limited.
It should be noted that, the steps of adding RMSE calculation and parameter adjustment can further optimize the intelligent ammeter operation error measurement model, and improve the accuracy and reliability of error measurement, thereby better meeting the actual application requirements.
Compared with the prior art, the invention has at least the following beneficial technical effects:
(1) According to the invention, even if the number of the intelligent electric meters is continuously increased, the topology structure of the intelligent electric meters is changed, the topology structure of the intelligent electric meters in the platform area can be obtained in real time through the graph neural network, and the topology structure of the intelligent electric meters is transmitted through the message transmission neural network, so that the operation error measurement model of the intelligent electric meters is automatically updated, a tester is not required to reset equation set parameters and newly establish a new linear equation set, and the error measurement efficiency is improved.
(2) According to the invention, based on a sparse optimization mode, the solution of the linear equation set can be thinned by introducing prior information, namely, a plurality of variables in the solution are zero, and the original large-scale linear equation set problem can be converted into a smaller-scale optimization problem, so that only a few variables need to be calculated in the process of solving the intelligent ammeter operation error measurement equation set, the calculation complexity and the difficulty of solving are greatly reduced, and the error measurement efficiency is improved.
(3) In the invention, based on a sparse optimization mode, the algorithm can be promoted to calculate a sparser solution through regularization constraint, so that the calculation of an invalid solution is avoided, and the accuracy of error measurement is improved.
Example 2
In one embodiment, the present invention provides a smart meter operation error measurement system for executing the smart meter operation error measurement method in embodiment 1.
The intelligent ammeter operation error measurement system provided by the invention can realize the steps and effects of the intelligent ammeter operation error measurement method in the embodiment 1, and in order to avoid repetition, the invention is not repeated.
Compared with the prior art, the invention has at least the following beneficial technical effects:
(1) According to the invention, even if the number of the intelligent electric meters is continuously increased, the topology structure of the intelligent electric meters is changed, the topology structure of the intelligent electric meters in the platform area can be obtained in real time through the graph neural network, and the topology structure of the intelligent electric meters is transmitted through the message transmission neural network, so that the operation error measurement model of the intelligent electric meters is automatically updated, a tester is not required to reset equation set parameters and newly establish a new linear equation set, and the error measurement efficiency is improved.
(2) According to the invention, based on a sparse optimization mode, the solution of the linear equation set can be thinned by introducing prior information, namely, a plurality of variables in the solution are zero, and the original large-scale linear equation set problem can be converted into a smaller-scale optimization problem, so that only a few variables need to be calculated in the process of solving the intelligent ammeter operation error measurement equation set, the calculation complexity and the difficulty of solving are greatly reduced, and the error measurement efficiency is improved.
(3) In the invention, based on a sparse optimization mode, the algorithm can be promoted to calculate a sparser solution through regularization constraint, so that the calculation of an invalid solution is avoided, and the accuracy of error measurement is improved.
The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.
The foregoing examples illustrate only a few embodiments of the invention, which are described in detail and are not to be construed as limiting the scope of the invention. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the invention, which are all within the scope of the invention. Accordingly, the scope of protection of the present invention is to be determined by the appended claims.
Claims (7)
1. The utility model provides a smart electric meter operation error measurement method which is characterized in that the method comprises the following steps:
s101: acquiring a smart meter topological structure of a platform region through a graph neural network, and transmitting the smart meter topological structure through a message transmission neural network, wherein the smart meter topological structure comprises a plurality of summary tables and a plurality of user meters connected with the summary tables;
s102: according to the intelligent ammeter topological structure, an intelligent ammeter operation error measurement model is constructed based on an energy conservation law:
wherein y (i) represents the measured value of the i-th periodic total table, x j (i) A measurement value representing the jth user meter of the ith period, n representing the number of user meters, ε j Representing the relative error of the jth user meter, E (i) representing the energy loss of the ith period, said energy loss being calculated from the line loss rate;
s103: acquiring intelligent ammeter data of a plurality of periods, and preprocessing the intelligent ammeter data to remove abnormal data;
s104: establishing a smart meter operation error measurement equation set according to the preprocessed smart meter data;
s105: solving the intelligent ammeter operation error measurement equation set based on a sparse optimization mode to obtain the relative error and the line loss rate of each user ammeter;
s106: when the relative error of the user ammeter is larger than a preset error value or the line loss rate is in a preset range, determining that the user ammeter has faults, and giving an alarm;
wherein, the step S104 specifically includes:
the intelligent ammeter operation error measurement equation set is established through the following formula:
wherein ,aj Representing the error coefficient of the jth user meter, b 0 Represents the line loss rate, w 0 Representing the fixed loss x j (i) Representing a measured value of a jth user meter in ith smart meter data, y (i) representing a measured value of a total table in ith smart meter data, Δy (i) representing a total power consumption in ith smart meter data, i=1, 2, …, m, m representing a total number of smart meter data, j=1, 2, …, n, n representing the number of user meters;
wherein the error coefficient a j The expression of (2) is:
wherein ,εj Representing the relative error of the jth user meter;
wherein, the step S105 specifically includes:
s1051: constructing a sparse optimization model of the intelligent ammeter operation error measurement equation set:
A＝[a 1 …a j ],
W＝[w 0 …w 0 ],ΔY＝[Δy(1)…Δy(m)],Y＝[y(1)…y(m)]
wherein A represents an error coefficient vector, X represents a measured value matrix of the user electricity meter, W represents a fixed loss matrix, deltaY represents a total power consumption vector, Y represents a total measured value vector, lambda represents a regularized parameter of relative error, and mu represents a lineThe regularization parameters of the loss are set, I.I 1 A norm is represented by a number of norms, I.I 2 Representing the two norms, y 0 A priori term representing line loss;
s1052: fixing the line loss rate, and updating a sparse optimization model of the intelligent ammeter operation error measurement equation set:
s1053: order theg(Z)＝λ||Z|| 1 An augmented lagrangian function is constructed:
wherein ρ represents a penalty parameter, and Q represents a penalty variable;
s1054: updating X, Z and Q according to the augmented Lagrangian function, and calculating f (X);
s1055: according toCalculating an error coefficient vector A;
s1056: according to the error coefficient vector A andcalculating the relative error epsilon of the jth user ammeter j ；
Wherein, the S105 further includes:
s1057: fixing the relative error, updating a sparse optimization model of the intelligent ammeter operation error measurement equation set, solving the line loss rate, wherein the updated sparse optimization model is as follows:
s1058: order theObtaining a second derivative h' (delta Y) of h (delta Y);
s1059: let h "(Δy) =0, calculate the line loss rate b 0 。
2. The method for measuring the operation error of the smart meter according to claim 1, wherein the obtaining the topology of the smart meter of the area through the graph neural network specifically comprises:
acquiring a connection relation between a summary list and the user electric meters through a graph neural network;
according to the connection relation between the summary list and the user electric meters and the connection relation between the user electric meters, each intelligent electric meter is taken as a node, and each connection relation is taken as an edge, so that intelligent electric meter graph data are constructed;
and converting nodes and edges in the intelligent ammeter graph data into low-dimensional vectors through a GCN graph embedding model to obtain embedded representation of the intelligent ammeter topological structure.
3. The smart meter operation error measurement method of claim 1, wherein the messaging neural network comprises a plurality of nodes, and the delivering the smart meter topology through the messaging neural network specifically comprises:
acquiring hidden states of nodes i adjacent to each node v after t times of message exchangeAnd the original hidden state of each node v +.>
By message transfer functionHiding state of adjacent node i after t times of message exchangeAnd the original hidden state of each node v +.>Combining:
where m represents a message transfer function, b 1 、b 2 Representing the offset;
according to the message transfer function, obtaining the message exchanged by the node v at t+1 times, and transferring:
wherein ,a message indicating the t+1st exchange of the node v, N (v) indicating the total number of nodes;
updating the message transfer function based on previous hidden statesAnd aggregating the messages to calculate a new hidden state +/for each node>Repeating the above process, and iterating until reaching the preset iteration times.
4. The smart meter operation error measurement method according to claim 1, wherein the energy loss E (i) of the i-th cycle is specifically:
wherein ,eM (i) Represents the self-loss of the electric energy meter in the ith period, e N (i) Represents the ith periodic leakage loss, e L (i) Indicating the i-th periodic line loss,indicating the rated power of the jth consumer meter, t (i) indicating the meter run time of the ith period, U indicating the line voltage, sigma indicating the line leakage conductivity, +.>Representing the line current of the jth user ammeter in the ith period, R j The line resistance from the jth user ammeter to the summary table is represented, and n represents the number of the user ammeter;
self-loss e of electric energy meter which is kept unchanged in each period M (i) Defined as fixed loss w 0 ；
Leakage loss e which will be difficult to accurately measure at each cycle N (i) And line loss e L (i) Defining as a change loss, and establishing a relationship between the change loss and a measurement value of a total table through a line loss rate:
e N (i)+e L (i)＝b 0 ·Δy(i)＝b 0 ·[y(i)-y(i-1)]；
wherein ,b0 The line loss rate is represented, and Δy (i) represents the total power consumption in the ith cycle.
5. The smart meter operation error measurement method as claimed in claim 4, wherein S103 specifically comprises:
s1031: comparing the sum of the measured values of all the user electric meters in the intelligent electric meter data with the measured value of the total table;
s1032: retaining the data of the measured values of the user electric meters, wherein the sum of the measured values of the user electric meters is smaller than or equal to the sum of the measured values of the total table, and removing the data of the measured values of the user electric meters, wherein the sum of the measured values of the user electric meters is larger than the sum of the measured values of the total table;
s1033: calculating the energy loss rate of each intelligent ammeter data:
wherein ,γi Energy loss rate of smart meter data representing the ith cycle, y (i) represents a measurement value of the total table of the ith cycle, x j (i) A measurement value representing the jth meter of the ith period, j=1, 2, …, n, n representing the number of meters;
s1034: and reserving the intelligent ammeter data with the energy loss rate smaller than or equal to the preset loss rate, and removing the intelligent ammeter data with the energy loss rate larger than the preset loss rate.
6. The smart meter operation error measurement method of claim 1, further comprising:
s107: calculating Root Mean Square Error (RMSE) of the intelligent ammeter operation error measurement model:
wherein ,εj Representing the relative error of the jth user ammeter calculated by the intelligent ammeter operation error measurement model,representing the actual error of the jth user meter;
s108: when the root mean square error of the intelligent ammeter operation error measurement model is larger than an error threshold value, adjusting regularization parameters lambda of relative errors, regularization parameters mu of line losses and punishment parameters rho, and re-calculating.
7. A smart meter operation error measurement system for performing the smart meter operation error measurement method of any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310987000.4A CN116718979B (en) | 2023-08-08 | 2023-08-08 | Smart electric meter operation error measurement method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310987000.4A CN116718979B (en) | 2023-08-08 | 2023-08-08 | Smart electric meter operation error measurement method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116718979A CN116718979A (en) | 2023-09-08 |
| CN116718979B true CN116718979B (en) | 2023-10-24 |
Family
ID=87866430
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310987000.4A Active CN116718979B (en) | 2023-08-08 | 2023-08-08 | Smart electric meter operation error measurement method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116718979B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2007063180A1 (en) * | 2005-12-01 | 2007-06-07 | Valtion Teknillinen Tutkimuskeskus | Method and system for the calibration of meters |
| CN112486959A (en) * | 2020-11-03 | 2021-03-12 | 国网河南省电力公司郑州供电公司 | Graph database technology-based intelligent identification method for dynamic topological structure of power grid |
| CN112819682A (en) * | 2019-11-15 | 2021-05-18 | 英特尔公司 | Shrinking arithmetic of sparse data |
| CN113032795A (en) * | 2019-12-24 | 2021-06-25 | 图灵人工智能研究院（南京）有限公司 | Data processing method, system and storage medium for electricity consumption data |
| CN113281697A (en) * | 2021-05-20 | 2021-08-20 | 国网河南省电力公司营销服务中心 | Operation error online analysis method and system |
| WO2022110558A1 (en) * | 2020-11-25 | 2022-06-02 | 国网湖南省电力有限公司 | Smart electricity meter malfunction early warning method and device |
| CN115561699A (en) * | 2022-12-05 | 2023-01-03 | 北京志翔科技股份有限公司 | Operation error estimation method and device |
| CN115877312A (en) * | 2022-10-26 | 2023-03-31 | 国家电网有限公司 | Electric energy meter informatization evaluation calibration model based on station area electric energy conservation |
-
2023
- 2023-08-08 CN CN202310987000.4A patent/CN116718979B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2007063180A1 (en) * | 2005-12-01 | 2007-06-07 | Valtion Teknillinen Tutkimuskeskus | Method and system for the calibration of meters |
| CN112819682A (en) * | 2019-11-15 | 2021-05-18 | 英特尔公司 | Shrinking arithmetic of sparse data |
| CN113032795A (en) * | 2019-12-24 | 2021-06-25 | 图灵人工智能研究院（南京）有限公司 | Data processing method, system and storage medium for electricity consumption data |
| CN112486959A (en) * | 2020-11-03 | 2021-03-12 | 国网河南省电力公司郑州供电公司 | Graph database technology-based intelligent identification method for dynamic topological structure of power grid |
| WO2022110558A1 (en) * | 2020-11-25 | 2022-06-02 | 国网湖南省电力有限公司 | Smart electricity meter malfunction early warning method and device |
| CN113281697A (en) * | 2021-05-20 | 2021-08-20 | 国网河南省电力公司营销服务中心 | Operation error online analysis method and system |
| CN115877312A (en) * | 2022-10-26 | 2023-03-31 | 国家电网有限公司 | Electric energy meter informatization evaluation calibration model based on station area electric energy conservation |
| CN115561699A (en) * | 2022-12-05 | 2023-01-03 | 北京志翔科技股份有限公司 | Operation error estimation method and device |
Non-Patent Citations (1)
| Title |
|---|
| 基于稀疏优化的智能电表误差估计方法;吕玉玲 等;重庆师范大学学报(自然科学版);第第40卷卷(第第3期期);1-7 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116718979A (en) | 2023-09-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109459669B (en) | 10kV power distribution network single-phase earth fault section positioning method | |
| CN107666149B (en) | Medium-voltage distribution network line loss calculation method | |
| CN113312413B (en) | Enterprise carbon emission abnormity monitoring method based on electric power big data | |
| WO2022021726A1 (en) | Pmu-based power system state estimation performance evaluation method | |
| EP2653835B1 (en) | A method and system for optimizing a composite load disaggregation | |
| CN108306303A (en) | Voltage stability evaluation method considering load increase and new energy output randomness | |
| Li et al. | Ensemble offshore wind turbine power curve modelling–an integration of isolation forest, fast radial basis function neural network, and metaheuristic algorithm | |
| CN111625991A (en) | Low-voltage distribution network topology verification method | |
| CN113687176B (en) | Deep neural network-based power consumption abnormity detection method and system | |
| CN112149873A (en) | Low-voltage transformer area line loss reasonable interval prediction method based on deep learning | |
| CN107133721A (en) | Power distribution network based on gray theory measures bad data relation factor analysis method | |
| CN116148753A (en) | Intelligent electric energy meter operation error monitoring system | |
| CN108520310A (en) | Wind speed forecasting method based on G-L mixed noise characteristic v- support vector regressions | |
| CN116796644A (en) | Wind farm parameter identification method based on multi-agent SAC deep reinforcement learning | |
| CN107257130A (en) | The low-voltage network loss computing method of decoupling is measured based on region | |
| CN113612237A (en) | Method for positioning resonance-induced subsynchronous oscillation source in offshore wind farm | |
| CN106372440B (en) | A kind of adaptive robust state estimation method of the power distribution network of parallel computation and device | |
| CN108564112B (en) | Unified fault positioning method for comprehensive energy system | |
| Wang et al. | Probabilistic power curve estimation based on meteorological factors and density LSTM | |
| CN116718979B (en) | Smart electric meter operation error measurement method and system | |
| CN117421622A (en) | Low-voltage transformer area line loss calculation method and system based on clustering algorithm | |
| CN111091141A (en) | Photovoltaic backboard fault diagnosis method based on layered Softmax | |
| CN116184303A (en) | Electric energy meter error online monitoring method and system based on layering and grading | |
| CN108761258A (en) | Transformer short period overload capability assessment system based on artificial intelligence and big data technology | |
| CN112241812A (en) | Low-voltage distribution network topology identification method based on unilateral optimization and genetic algorithm cooperation |
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 |