Online estimation method and system suitable for cluster type intelligent electric meter error analysis

Abstract

The invention discloses an on-line estimation method and system suitable for cluster type intelligent electric meter error analysis, and the method comprises the steps: building an intelligent electric meter error on-line calculation model through an energy conservation principle based on the characteristic that most intelligent electric meters are distributed according to a tree topology structure in reality, and solving the model to obtain an error estimation value. On this basis, non-uniformity of error information contained in measurement data in time is considered, a weight coefficient represented by attenuation factors is introduced to construct a loss function of a calculation model, different attenuation factors are adopted according to the condition number of a coefficient matrix to improve the tracking capability of an algorithm, an error estimation value is obtained by deriving the loss function, and the tracking precision of the algorithm is improved. And the error estimation value of the electric energy meter is updated in real time through newly collected data, so that the solving precision of the error estimation value is improved.

Description

Technical Field

[0001] The present application relates to the technical field of electric energy metering in power systems, and in particular to an online estimation method and system suitable for error analysis of clustered smart meters. Background Art

[0002] my country is accelerating its energy transformation and promoting the construction of a new power system with new energy as the main body. However, with the high proportion of new energy with large volatility, strong randomness and high prediction difficulty connected to the grid, as well as the high proportion of access to multiple data sources and multiple types of power equipment, a "grid" with many devices, complex data and high difficulty in operation and maintenance technology has been formed, which poses a huge challenge to the power balance, safe power supply and operation and maintenance detection of the power grid. In this case, the traditional operation and maintenance methods can no longer meet the high-speed and high-intensity development needs of the power grid. Digital twin technology, which is supported by big data and fully relies on emerging technologies such as cloud computing, blockchain and artificial intelligence, can fully empower smart grids and smart power plants, realize full coverage and real-time monitoring of the entire process of power production and grid operation and maintenance, and make the grid "objective, controllable and measurable".

[0003] However, the construction of the digital twin system in the power industry requires a large amount of real-time measurement data, which puts forward the requirements of scale, high timeliness and high accuracy for the metering technology. For the low-voltage area in the new power system, the main metering equipment is the smart meter. However, the current measurement data has poor real-time performance and low accuracy, which makes it difficult to meet the needs of the construction of the digital twin system. With the continuous development of big data technology and the continuous improvement of the smart grid information collection system, the online metering technology of smart electric energy meters based on data analysis provides a solution to the above problems. Online metering of electric energy meters is a new metering technology that uses data analysis algorithms to analyze the measurement data of electric energy meters and estimate the errors of electric energy meters. Compared with the one-to-one detection method of traditional laboratories, online metering can estimate the error values ​​of multiple electric energy meters in real time, and has the characteristics of high detection efficiency and strong real-time performance. Regulatory departments can use the online metering technology of smart electric energy meters to monitor the operating status of batch electric energy meters online and solve many problems in the current regulatory process. Therefore, online metering of smart electric energy meters is a technology with good development prospects and great practical value.

[0004] The current online error technology and online error estimation algorithm of smart electricity meters require screening of the collected data, and lack real-time performance; the error matrix composed of electricity measurement data has serious pathological performance, and when the number of electricity meters in the system is large, the accuracy of error estimation decreases. Summary of the invention

[0005] In order to solve the problems of insufficient real-time and accuracy in the existing online error estimation technology for smart meters, the present application proposes an online estimation method and system suitable for error analysis of clustered smart meters. Based on the characteristic that smart meters are mostly distributed in a tree-like topology structure in reality, the present application constructs an online calculation model of smart meter errors through the principle of conservation of energy. The error estimate can be obtained by solving this model. On this basis, the temporal inhomogeneity of the error information contained in the measurement data is considered, and the weight coefficient represented by the attenuation factor is introduced to construct the loss function of the calculation model. Different attenuation factors are used according to the condition number of the coefficient matrix to improve the tracking ability of the algorithm. The error estimate is obtained by derivatizing the loss function, and the error estimate of the electric energy meter is updated in real time through the newly collected data, thereby improving the accuracy of solving the error estimate.

[0006] This application is implemented through the following technical solutions:

[0007] An online estimation method suitable for error analysis of clustered smart meters, the online estimation method comprising:

[0008] According to the topological structure of the low-voltage area and based on the law of conservation of energy, an online calculation model for smart meter errors is established;

[0009] According to the temporal non-uniformity of the error information contained in the measurement data, a weight coefficient represented by an attenuation factor is introduced to construct a loss function of the online calculation model of the smart meter error, and an error estimation value is obtained by deriving the loss function;

[0010] Adopting a recursive error estimation algorithm to update the smart meter error estimation value in real time;

[0011] According to the condition number of the coefficient matrix of the online calculation model of the smart meter error, a variable attenuation factor is introduced into the weight coefficient to correct the error estimation algorithm.

[0012] In some implementations, the online calculation model for smart meter errors is established based on the topological structure of the low-voltage area and the law of conservation of energy, specifically including:

[0013] According to the tree topology of the low-voltage area smart meter, the energy conservation relationship of the system is obtained as follows: the electric energy value flowing into the main meter is equal to the sum of the electric energy values ​​flowing into each sub-meter under the main meter;

[0014] The energy conservation relation considering the measurement errors of the total meter and sub-meters is obtained;

[0015] The total meter is used as the calibration meter, that is, the measurement error of the total meter is regarded as zero, and the updated energy conservation relationship is obtained;

[0016] Based on the measurement data of the total table and the sub-tables collected multiple times, multiple equation groups are established according to the updated energy conservation relationship; wherein the number of times the measurement data is collected is greater than or equal to the number of the sub-tables;

[0017] The least square method is used to solve the equations, and the online calculation model of the smart meter error is expressed as:

[0018] λ=(E T E) -1 E T α

[0019] Where λ is the error-related parameter matrix of each smart meter; E is the electric energy measurement value matrix of several sub-meters, which serves as the coefficient matrix of the model; α is the total meter electric energy measurement value matrix.

[0020] In some implementations, the loss function of the online calculation model of the smart meter error is constructed by introducing a weight coefficient represented by an attenuation factor according to the temporal inhomogeneity of the error information contained in the measurement data, and the error estimation value is obtained by deriving the loss function, specifically including:

[0021] Assume that the weight of the measurement data at the i-th moment is ω i , the loss function of solving the equation is expressed as:

[0022]

[0023] Where e(i) is the residual value at the i-th moment, ω is the weight matrix, and e is the residual matrix.

[0024] Deriving the loss function and setting the derivative to zero minimizes the loss function, the error estimate is:

[0025] λ ω =(E T ωE) -1 E T ωα.

[0026] In some embodiments, an exponential function is used to represent the weight, Where μ is the attenuation factor, n s is the number of measurement data samples.

[0027] In some implementations, the recursive error estimation algorithm is specifically expressed as:

[0028]

[0029] In the formula, λ ω (n s +1) is the error estimate at the current moment, λ ω (n s) is the error estimate at the previous moment, K and H are the intermediate matrices used for calculation, and I is the unit matrix.

[0030] In some implementations, the condition number of the coefficient matrix of the online calculation model of the smart meter error is used to introduce a variable attenuation factor into the weight coefficient, specifically:

[0031] Let the attenuation factor μ a =|cond(E)γ / (1-cond(E)γ)|, then the attenuation factor is obtained according to the condition number:

[0032]

[0033] Among them, γ is the relative error of the coefficient matrix; μ min is the minimum value of μ; cond(E) is the condition number of the coefficient matrix, and n is the number of sub-tables.

[0034] In some implementations, after introducing a variable attenuation factor, the error estimation algorithm is expressed as:

[0035]

[0036] In the formula, λ ω (n s +1) is the error estimate at the current moment, λ ω (n s ) is the error estimate of the previous moment, μ(n s ) is the attenuation factor at the previous moment, K and H are the intermediate matrices used for calculation, and I is the unit matrix.

[0037] In a second aspect, the present application proposes an online estimation system suitable for cluster smart meter error analysis, the online estimation system comprising:

[0038] A model building module, wherein the model building module establishes an online calculation model of smart meter errors according to the topological structure of the low-voltage area and based on the law of conservation of energy;

[0039] An error estimation module, wherein the error estimation module introduces a weight coefficient represented by an attenuation factor to construct a loss function of the online calculation model of the smart meter error according to the temporal inhomogeneity of the error information contained in the measurement data, and obtains an error estimation value of the smart meter by deriving the loss function;

[0040] A recursive solution module, wherein the recursive solution module adopts a recursive error estimation algorithm to update the error estimation value of the smart meter in real time;

[0041] And, a correction module, wherein the correction module introduces a variable attenuation factor into the weight coefficient according to the condition number of the coefficient matrix of the online calculation model of the smart meter error, and corrects the error estimation algorithm.

[0042] In a third aspect, the present application proposes an electronic device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0043] In a fourth aspect, the present application proposes a computer-readable storage medium having a computer program stored thereon, wherein the computer program implements the steps of the above method when executed by a processor.

[0044] The present application proposes an online estimation method and system for cluster smart meter error analysis. Starting from the topological structure of the cluster smart meter, an online error calculation model is constructed based on the law of conservation of energy. The traditional solution method believes that the influence of measurement data at different times on the error is consistent, but in actual situations, the intervals for measuring data collection are large, and recent measurement data may contain more current period electric energy meter error information than long-term measurement data. Therefore, the present application assigns corresponding weights to measurement data at different times, reduces the influence of early data or unreasonable data on error estimation, and improves the accuracy of error estimation.

[0045] The present application proposes an online estimation method and system suitable for error analysis of clustered smart meters. A recursive error estimation algorithm is used to solve the error estimation value, and an exponential decay factor is introduced to characterize the influence of measurement data at different times on the error estimation. At the same time, in order to reduce the influence of the morbidity of the coefficient matrix of the calculation model on the error estimation solution, the attenuation factor is selected by analyzing the condition number of the coefficient matrix, which improves the solution accuracy and tracking ability of the algorithm, provides technical support for the online measurement of smart meters, and helps to promote the application of digital twin technology in power systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The drawings described herein are used to provide a further understanding of the embodiments of the present application, constitute a part of the present application, and do not constitute a limitation on the embodiments of the present application. In the drawings:

[0047] Figure 1 A flow chart of the method proposed in the embodiment of the present application;

[0048] Figure 2 An example of a low-voltage area topology structure provided in an embodiment of the present application;

[0049] Figure 3 A block diagram of the system principles provided in an embodiment of the present application. DETAILED DESCRIPTION

[0050] In order to make the objectives, technical solutions and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with examples and drawings. The illustrative implementation scheme of the present application and its description are only used to explain the present application and are not intended to limit the present application.

[0051] Example:

[0052] The existing online error estimation technology for smart meters has problems such as insufficient real-time performance and low error estimation accuracy. To address this, this embodiment proposes an online estimation method suitable for cluster smart meter error analysis.

[0053] like Figure 1 As shown, the method proposed in this embodiment specifically includes the following steps:

[0054] Step 100, establishing an online calculation model for smart meter errors according to the topological structure of the low-voltage area and based on the law of conservation of energy.

[0055] like Figure 2 As shown in the figure, the connection mode of the smart meters in the low-voltage area is simplified to form a cluster that is usually a tree topology. In the low-voltage area, the user's smart meter is installed on the secondary side of the distribution transformer. The smart meters are distributed according to the topological structure. Each area will be installed with a transformer, a smart meter master meter and several users' smart meters. The smart meter master meter is connected to the transformer on one side and two user-end sub-meters on the other side. The master meter records the electricity consumption of the entire area, and the sub-meter records the electricity consumption of a single user. The entire low-voltage area is regarded as a generalized node. Under ideal conditions, the electric energy value flowing into the master meter is equal to the sum of the electric energy values ​​flowing into the sub-meter. Let the master meter be M0, and there are n sub-meters under the master meter, represented by M j (j=1,2,...,n). The energy conservation relationship in the system can be expressed as:

[0056]

[0057] Where W j (i) represents the actual amount of electricity flowing through the j-th meter in the i-th time period; W0(i) represents the actual amount of electricity flowing through the i-th time period.

[0058] During the operation of the electric energy meter, there is a measurement error. After considering the error factor, formula (1) can be expressed as:

[0059]

[0060] Where E j (i) represents the electric energy measurement value of the jth electric meter in the i-th period; E0(i) represents the electric energy measurement value of the total meter at the i-th time end; ε j It represents the measurement error of the jth electric meter. When j=0, ε0 is the measurement error of the total meter.

[0061] The total meter of smart meters is used as a calibration table for calibration and verification of electric energy in the power grid. The ordinary electric meters for residents are Class 2 electric energy meters, and the total meter is generally a Class 0.2 electric energy meter. The accuracy of the total meter is one level higher than that of the residential electric meter. The error of the total meter is small, and the impact on the error estimation can be ignored. The error of the total meter can be considered to be zero. At this time, formula (2) can be expressed as:

[0062]

[0063] In actual working conditions, it is necessary to consider the loss in the low-voltage power grid and the energy consumed in the line during the transmission process. The collected measurement data may contain abnormal values. After introducing loss and abnormal data, the energy conservation relationship can be expressed as:

[0064]

[0065] In the formula It represents the normal value of electric energy measurement of the jth electric meter in the i-th time period; Indicates the abnormal deviation of the energy measurement value of the jth meter in the i-th time period; when the data is normal

[0066] Let λ j =1 / (1+ε j ), then λ j is a parameter related to the meter error. By solving λ j The error estimation value of the jth meter can be obtained. Assuming that the current measurement data does not contain abnormal data and the line loss in the system is not considered, equation (4) is expressed as:

[0067]

[0068] Each time the data is collected, the relationship of formula (5) can be obtained. s (n s ≥n) times of measurement data, the obtained n s The equations are combined into a system of equations as follows:

[0069] Eλ=α (6)

[0070] Where:

[0071] The least square method can be used to solve equation (6):

[0072] λ=(E T E) -1 E T α (7)

[0073] Among them, λ is the error-related parameter matrix of each smart meter (sub-meter); E is the electric energy measurement value matrix of several sub-meters, which serves as the coefficient matrix of the model; α is the total meter electric energy measurement value matrix.

[0074] Solving equation (7) yields λ j , then the error of the energy meter can be solved as:

[0075]

[0076] Step 200, according to the temporal inhomogeneity of the error information contained in the measurement data, a weight coefficient represented by an attenuation factor is introduced to construct a loss function of the calculation model, and an error estimation value is obtained by deriving the loss function.

[0077] When using the least squares method to solve the electric energy metering error, the data at each measurement time is considered to have the same weight. However, in some cases, the time interval for collecting measurement data is large, and recent measurement data may contain more error information of the current meter than long-term measurement data, and is more important for error estimation, and can better reflect the characteristics of the recent error of the meter. Therefore, this embodiment introduces corresponding weights to the measurement data collected at different times to reduce the impact of early data or unreasonable data on error estimation.

[0078] Assume that the weight of the measurement data at the i-th moment is ω i , the loss function of solving the equation is expressed as:

[0079]

[0080] Where e(i) is the residual value at the i-th moment, ω is the weight matrix, and e is the residual matrix.

[0081] The weight coefficient of the measured data is a non-negative value. For the derivative in equation (9), setting the derivative to zero minimizes the loss function and the error estimate is:

[0082] λ ω =(E T ωE) -1 E T ωα (10)

[0083] When the weight matrix is ​​the unit matrix ω = I, the weight of the data at each moment is equal and is 1, which is the least squares method. When the weight coefficients are different, the larger the weight, the more important the data is, and the greater its importance to the error estimation.

[0084] Step 300: using a recursive error estimation algorithm to update the smart meter error estimation value in real time.

[0085] The online estimation of meter error has high requirements for real-time performance. After the new measurement data is collected, the error estimation value needs to be updated in real time. This embodiment adopts a recursive error estimation method to update the error estimation value in real time. Specifically:

[0086] When there is n s When measuring the data, the error estimate is:

[0087] λ ω (n s )=(E T (n s )ω(n s )E(n s )) -1 E T (n s )ω(n s )α(n s ) (11)

[0088] Let H(n s )=(E T (n s )ω(n s )E(n s )) -1 , then:

[0089] λ ω (n s )=H(n s )E T (n s )ω(n s )α(n s ) (12)

[0090] When a new measurement data is added, the number of data samples becomes n s +1, and its error estimate can be expressed as:

[0091] λ ω (n s +1)=H(n s +1)E T (n s +1)ω(n s +1)α(n s +1) (13)

[0092] In order to reduce the amount of calculation and facilitate subsequent analysis, an exponential function is used to represent the weights: μ is the attenuation factor, At this time, we can get:

[0093]

[0094] According to the matrix inversion formula:

[0095] (A -1 +BCD) -1 =A-AB(DAB+C) -1 DA (15)

[0096] Substitute A=H(n s ) / μ,B=E(n s +1), C=I, D=E T (n s +1), we get:

[0097]

[0098] in, Is a constant.

[0099] Formula (13) can be expressed as:

[0100]

[0101] From formula (12), we can get:

[0102] E T (n s )ω(n s )E(n s )=H -1 (n s ) ω (n s )(18)

[0103] Substituting formula (18) into formula (17), we get:

[0104] λ ω (n s +1)=H(n s +1)[μH -1 (n s ) ω (n s )+E(n s +1)α(n s +1)] (19)

[0105] According to formula (14), we have:

[0106]

[0107] Substituting formula (20) into formula (19), the simplified calculation yields:

[0108] λ ω (n s +1)=λ ω (n s )+H(ns +1)E(n s +1)[α(n s +1)-E T (n s +1)λ ω (n s )] (twenty one)

[0109] Let K(n s )=H(n s +1)E(n s +1) into formula (16) to obtain:

[0110]

[0111] Formula (16) can be expressed as:

[0112]

[0113] In summary, the error estimate can be expressed as:

[0114]

[0115] K and H are intermediate matrices used for calculation, and their respective expressions are recorded above; I is the unit matrix; “|” indicates merging with the matrix on the left into a new matrix.

[0116] It can be seen from the above formula that the recursive error estimation process does not involve matrix inverse calculation and is only related to the parameters at the previous moment. Therefore, this embodiment can improve the real-time performance, solution accuracy and applicability of the algorithm.

[0117] Step 400, the condition number of the model coefficient matrix is ​​calculated online according to the error, a variable attenuation factor is introduced into the weight coefficient, and the error estimation algorithm is corrected.

[0118] Since the weight coefficient is represented by an exponential function, as the number of information collection increases, if the attenuation factor is too large, it will affect the tracking ability of the algorithm, and if it is too small, it will affect the accuracy of the algorithm. Therefore, in order to solve the contradictory relationship between the accuracy of the algorithm and the tracking ability, a variable attenuation factor is introduced into the weight coefficient to correct the calculation model of the error estimate. The condition number of the matrix can be used to measure the state of the matrix. The larger the condition number of the matrix, the closer the matrix is ​​to a singular matrix, and the matrix is ​​an ill-conditioned matrix. When the ill-conditioned matrix is ​​used as a coefficient matrix to solve the equation, the more serious the matrix ill-conditioning is, and the greater the error of the model solution is.

[0119] When the coefficient matrix E is a square matrix, and there are deviations δE and δα between E and α respectively, when equation (6) satisfies (E+δE)(λ+δλ)=(α+δα), the error of the condition number in solving the equation system is:

[0120]

[0121] When the relative errors of E and α are fixed, ||δE|| / ||E|| and ||δα|| / ||α||, the larger the condition number cond(E), the larger the relative error of the solution ||δλ|| / ||λ||, the more unstable the solution is, and the accuracy is reduced. Since the condition number of the matrix can reflect the data status of the matrix itself and affect the accuracy of the solution of the equation, the attenuation factor is designed according to the condition number. Each time the error is estimated, a different attenuation factor is used according to the coefficient matrix to improve the accuracy of the algorithm while maintaining a high tracking ability.

[0122] In formula (25), α is the total meter electric energy measurement value. In the online error calculation model, the total meter electric energy measurement value can be considered as the standard value, and there is no error, so ||δα|| is 0. Since the maximum value of the normal meter measurement error is 2%, the maximum value of ||δE|| / ||E|| is 2%, and the value range is [0, 0.02]. Let γ = ||δE|| / ||E||, then the attenuation factor μ can be set a =|cond(E)γ / (1-cond(E)γ)|, then the attenuation factor is obtained according to the condition number:

[0123]

[0124] Where μ min is the minimum value of μ. When there are n sub-meters in the system, its possible values ​​are:

[0125]

[0126] γ is a custom value in the interval [0,0.02]. When the coefficient matrix is ​​ill-conditioned, the condition number cond(E) is large, and μ a When the condition number of the matrix is ​​small, the influence of the matrix on the solution of the equation is small, μ a It can be a value greater than 1. In this case, select μ min , which can improve the tracking ability of the algorithm.

[0127] After introducing the variable attenuation factor, the error estimate is:

[0128]

[0129] By solving equation (28), online estimation of electric energy error can be achieved.

[0130] The online estimation method for smart meter errors proposed in this embodiment is based on the topological structure of smart meters in reality. An online calculation model for smart meter errors is constructed through the principle of conservation of energy. Error estimates can be obtained by solving this model. On this basis, the temporal non-uniformity of error information contained in the measurement data (measurement data) is considered, and a weight coefficient represented by an attenuation factor is introduced to construct a loss function of the calculation model. A variable attenuation factor is used according to the condition number of the coefficient matrix to improve the tracking ability of the algorithm. The error estimate is obtained by deriving the loss function, and the meter error estimate is updated in real time through newly collected data, thereby improving the accuracy of solving the error estimate.

[0131] Based on the same technical concept as above, this embodiment also proposes an online estimation system suitable for cluster smart meter error analysis, such as Figure 3 As shown, the system proposed in this embodiment includes:

[0132] A model building module is provided, which establishes an online calculation model of smart meter errors according to the topological structure of the low-voltage area and based on the law of conservation of energy.

[0133] The error estimation module introduces a weight coefficient represented by an attenuation factor to construct a loss function of the calculation model according to the temporal inhomogeneity of the error information contained in the measurement data, and obtains an error estimation value by deriving the loss function.

[0134] The recursive solution module adopts a recursive error estimation algorithm to update the error estimation value of the smart meter in real time.

[0135] and a correction module, which calculates the condition number of the model coefficient matrix online according to the error, introduces a variable attenuation factor to the weight coefficient, and corrects the error estimation algorithm.

[0136] It should be noted that the specific implementation process of each functional module of the above system proposed in this embodiment is as described in steps 100 to 400 of the above method, and will not be repeated here.

[0137] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0138] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0139] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0140] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0141] The specific implementation methods described above further illustrate the purpose, technical solutions and beneficial effects of the present application in detail. It should be understood that the above description is only the specific implementation method of the present application and is not intended to limit the scope of protection of the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the scope of protection of the present application.

Claims

1. An online estimation method suitable for cluster smart meter error analysis, characterized in that: The online estimation method comprises: According to the topological structure of the low-voltage area and based on the law of conservation of energy, an online calculation model for smart meter errors is established; According to the temporal non-uniformity of the error information contained in the measurement data, a weight coefficient represented by an attenuation factor is introduced to construct a loss function of the online calculation model of the smart meter error, and an error estimation value is obtained by deriving the loss function; Adopting a recursive error estimation algorithm to update the smart meter error estimation value in real time; According to the condition number of the coefficient matrix of the online calculation model of the smart meter error, a variable attenuation factor is introduced into the weight coefficient to correct the error estimation algorithm.

2. The online estimation method for cluster smart meter error analysis according to claim 1, characterized in that: The online calculation model of smart meter error is established based on the topological structure of the low-voltage area and the law of conservation of energy, which specifically includes: According to the tree topology of the low-voltage area smart meter, the energy conservation relationship of the system is obtained as follows: the electric energy value flowing into the main meter is equal to the sum of the electric energy values ​​flowing into each sub-meter under the main meter; The energy conservation relation considering the measurement errors of the total meter and sub-meters is obtained; The total meter is used as the calibration meter, that is, the measurement error of the total meter is regarded as zero, and the updated energy conservation relationship is obtained; Based on the measurement data of the total table and the sub-tables collected multiple times, multiple equation groups are established according to the updated energy conservation relationship; wherein the number of times the measurement data is collected is greater than or equal to the number of the sub-tables; The least square method is used to solve the equations, and the online calculation model of the smart meter error is expressed as: λ=(E T E) -1 E T a Where λ is the error-related parameter matrix of each smart meter; E is the electric energy measurement value matrix of several sub-meters, which serves as the coefficient matrix of the model; α is the total meter electric energy measurement value matrix.

3. The online estimation method for cluster smart meter error analysis according to claim 2, characterized in that: The method of introducing a weight coefficient represented by an attenuation factor to construct a loss function of the online calculation model of the smart meter error according to the temporal inhomogeneity of the error information contained in the measurement data, and obtaining an error estimate by deriving the loss function, specifically includes: Assume that the weight of the measurement data at the i-th moment is ω i , the loss function of solving the equation is expressed as: Where e(i) is the residual value at the i-th moment, ω is the weight matrix, and e is the residual matrix. Deriving the loss function and setting the derivative to zero minimizes the loss function, the error estimate is: l ω =(E T ωE) -1 E T Oh.

4. The online estimation method for cluster smart meter error analysis according to claim 3 is characterized in that: The exponential function is used to represent the weight, Where μ is the attenuation factor, n s is the number of measurement data samples.

5. The online estimation method for cluster smart meter error analysis according to claim 4, characterized in that: The recursive error estimation algorithm is specifically expressed as: In the formula, λ ω (n s +1) is the error estimate at the current moment, λ ω (n s ) is the error estimate at the previous moment, K and H are the intermediate matrices used for calculation, and I is the unit matrix.

6. An online estimation method suitable for cluster smart meter error analysis according to any one of claims 1 to 5, characterized in that: The condition number of the coefficient matrix of the online calculation model of the smart meter error is used to introduce a variable attenuation factor into the weight coefficient, specifically: Let the attenuation factor μ a =|cond(E)γ / (1-cond(E)γ)|, then the attenuation factor is obtained according to the condition number: Among them, γ is the relative error of the coefficient matrix; μ min is the minimum value of μ; cond(E) is the condition number of the coefficient matrix, and n is the number of sub-tables.

7. The online estimation method and system for cluster smart meter error analysis according to claim 6, characterized in that: After introducing the variable attenuation factor, the error estimation algorithm is expressed as: In the formula, λ ω (n s +1) is the error estimate at the current moment, λ ω (n s ) is the error estimate of the previous moment, μ(n s ) is the attenuation factor at the previous moment, K and H are the intermediate matrices used for calculation, and I is the unit matrix.

8. An online estimation system suitable for cluster smart meter error analysis, characterized in that: The online estimation system comprises: A model building module, wherein the model building module establishes an online calculation model of smart meter errors according to the topological structure of the low-voltage area and based on the law of conservation of energy; An error estimation module, wherein the error estimation module introduces a weight coefficient represented by an attenuation factor to construct a loss function of the online calculation model of the smart meter error according to the temporal inhomogeneity of the error information contained in the measurement data, and obtains an error estimation value of the smart meter by deriving the loss function; A recursive solution module, wherein the recursive solution module adopts a recursive error estimation algorithm to update the error estimation value of the smart meter in real time; And, a correction module, wherein the correction module introduces a variable attenuation factor into the weight coefficient according to the condition number of the coefficient matrix of the online calculation model of the smart meter error, and corrects the error estimation algorithm.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images