An uncertainty evaluation method for an error online evaluation algorithm of an electric energy meter

Abstract

The application provides an uncertainty evaluation method of an electric energy meter error online evaluation algorithm, and belongs to the technical field of electric power metering and measurement. The method comprises the following steps: verifying the linear characteristics of the electric energy meter error online evaluation algorithm to be evaluated; for the electric energy meter error online evaluation algorithm that passes the verification, analyzing the uncertainty sources of the electric energy meter error online evaluation algorithm, and decomposing the uncertainty of the algorithm into multiple uncertainty components; for each uncertainty component, the GUM method or the Monte Carlo method is used for evaluation to obtain the evaluation results of each uncertainty component; and based on the evaluation results of each uncertainty component, the uncertainty of the electric energy meter error online evaluation algorithm is obtained. The method effectively solves the problem that the traditional analytic method cannot be used for comprehensive calculation, and overcomes the limitation of the Monte Carlo method in comprehensive analysis in each dimension.

Description

Technical Field

[0001] This invention belongs to the field of power metering and measurement technology, and specifically relates to an uncertainty assessment method for an online evaluation algorithm for electricity meter errors. Background Technology

[0002] When evaluating the performance of online error assessment algorithms for electricity meters, existing methods mainly rely on traditional error evaluation metrics, such as root mean square error (RMSE), mean absolute error (MAE), and mean absolute percentage error (MAPE). RMSE reflects the deviation between predicted and actual values ​​but is sensitive to outliers. MAE compensates for RMSE's shortcomings in handling outliers but still cannot fully reflect the algorithm's performance. MAPE represents the relative error between predicted and actual values, but since the true value may be zero, this metric cannot reflect the algorithm's performance and therefore cannot be calculated. Thus, traditional error evaluation metrics have limitations in algorithm performance evaluation; they can only provide an overall evaluation of the algorithm, cannot pinpoint the source of error, and cannot provide information about the algorithm's uncertainty.

[0003] To address the aforementioned issues, some studies have attempted to add other metrics to the evaluation methods. For example, some studies, in addition to using MAPE and RMSE, have added mean error (ME) as an evaluation metric to assess online error estimation algorithms. Other studies have used mean squared error (MSE) to evaluate the algorithms. Still other studies have used relative sum of squared errors (RSSE) for evaluation. Reference 4 uses false negative rate (FNR), false positive rate (FPR), and total positive rate (TPR) as evaluation metrics.

[0004] In uncertainty assessment, commonly used methods mainly include the Guide to the Expression of Uncertainty in Measurement (GUM) and the Monte Carlo method. While uncertainty assessment methods are widely applied across various fields, in-depth research is still lacking in uncertainty assessment for electricity meter error evaluation algorithms. The online evaluation algorithm for electricity meter errors has a complex model, not a simple linear model. This algorithm involves calculating the errors of hundreds of electricity meters in a distribution area, meaning the model needs to handle a large number of inputs and may introduce nonlinear factors during the actual solution process. Existing uncertainty assessment methods, such as the GUM method and the Monte Carlo method, cannot handle the complex relationships between multiple variables in the algorithm and the influence of multiple uncertainties in practical applications. Therefore, the traditional GUM method cannot directly and effectively assess the uncertainty of this algorithm.

[0005] Furthermore, the solution to the electricity meter error algorithm is affected by a variety of factors. For example, strong correlations between electricity consumption data, noise, quantization errors, line losses, and environmental factors can all affect the accuracy of the algorithm. If the Monte Carlo method is used for uncertainty assessment, it requires extensive simulation experiments for each influencing factor, which demands not only significant computational resources but also a considerable amount of time.

[0006] In summary, overcoming the limitations of purely analytical and Monte Carlo methods in analyzing algorithm uncertainty and comprehensively analyzing algorithm uncertainty has become an urgent technical problem to be solved. Summary of the Invention

[0007] The purpose of this invention is to solve the problems existing in the prior art and provide an uncertainty assessment method for an online evaluation algorithm for electricity meter errors. This method effectively solves the problem that traditional analytical methods cannot perform comprehensive calculations and also overcomes the limitations of the Monte Carlo method in comprehensive analysis across various dimensions.

[0008] This invention is achieved through the following technical solution:

[0009] In a first aspect, the present invention provides an uncertainty assessment method for an online evaluation algorithm for electricity meter errors, the uncertainty assessment method comprising:

[0010] The linearity of the online evaluation algorithm for the error of the electricity meter to be evaluated was verified;

[0011] For the verified online evaluation algorithm for electricity meter error, the sources of uncertainty in the algorithm are analyzed, and the uncertainty of the algorithm is decomposed into multiple uncertainty components.

[0012] For each uncertainty component, the GUM method or Monte Carlo method is used for evaluation to obtain the evaluation results of each uncertainty component;

[0013] Based on the evaluation results of each uncertainty component, the uncertainty of the online evaluation algorithm for electricity meter error is obtained.

[0014] A further improvement of the present invention is that:

[0015] The linearity of the online evaluation algorithm for the electricity meter error to be evaluated was verified, including:

[0016] Based on the topology of the online metering system for electricity meters in the low-voltage distribution area, an initial zero-error dataset is constructed, which includes the total meter electricity data and the sub-meter electricity data of each sub-meter.

[0017] Based on the zero-error dataset, the linearity of the online energy meter error evaluation algorithm is verified through linear experiments on the solution varying with the total meter error and linear experiments on the solution varying with the sub-meter error.

[0018] A further improvement of the present invention is that:

[0019] The linearity of the online energy meter error evaluation algorithm is verified by solving a linear experiment that varies with the total meter error. The steps include:

[0020] Step S410: Randomly set a first increment factor to adjust the total meter power data;

[0021] Step S420: Set the first random noise and load the sub-meter energy data and the adjusted total meter energy data into each sub-meter respectively;

[0022] Step S430: Solve the error online evaluation algorithm of the electricity meter, and obtain the first deviation value by subtracting the theoretical value from the actual value solved by the algorithm;

[0023] Step S440: Repeat steps S410-S430, and calculate the statistical results of the first deviation value;

[0024] Step S450: Based on the statistical results of the first deviation value, verify the linearity of the change in total meter error solved by the online evaluation algorithm for electricity meter error.

[0025] A further improvement of the present invention is that:

[0026] The linearity of the online energy meter error evaluation algorithm is verified through a linear experiment that solves the problem of how the meter error changes, including:

[0027] Step S510: Randomly set the second increment factor to adjust the sub-meter power data of each sub-meter;

[0028] Step S520: Set a second random noise and load it into the total meter energy data and the adjusted sub-meter energy data respectively;

[0029] Step S530: Solve the error online evaluation algorithm of the electricity meter, and subtract the theoretical value of the algorithm from the actual value obtained by the algorithm to obtain the second deviation value of each sub-meter;

[0030] Step S540: Repeat steps S510-S530, and calculate the statistical result of the second deviation value;

[0031] Step S550: Based on the statistical results of the second deviation value, verify the linear characteristics of the energy meter error online evaluation algorithm as a function of sub-meter error.

[0032] A further improvement of the present invention is that:

[0033] Multiple uncertainty components include the zero-point uncertainty component of the algorithm, the uncertainty component of the summary table, and the uncertainty component of the line loss calculation.

[0034] A further improvement of the present invention is that:

[0035] The zero-point uncertainty component is evaluated to obtain the zero-point uncertainty of the algorithm.

[0036] A further improvement of the present invention is that:

[0037] The steps involved in the evaluation include:

[0038] Step S310: Set the third random noise and load it into the main meter's energy data and the sub-meter's energy data for each sub-meter.

[0039] Step S320: Solve the calculation error of each sub-meter using the online evaluation algorithm for electricity meter error;

[0040] Step S330, repeat steps S310 and S320 to obtain multiple calculation errors for each sub-table, and calculate the mean and standard deviation of the multiple calculation errors for each sub-table.

[0041] Step S340: Calculate the uncertainty of each sub-table using the following formula, perform statistical analysis on the uncertainty of each sub-table, and obtain the zero-point uncertainty of the algorithm:

[0042]

[0043] Among them, u a The expression represents the uncertainty of the sub-table, a represents the mean, b represents the standard deviation, and max represents taking the larger of |a+2b| and |a-2b|.

[0044] A further improvement of the present invention is that:

[0045] Based on the evaluation results of each uncertainty component, the uncertainty of the online evaluation algorithm for electricity meter error is obtained, including:

[0046] Based on the evaluation results of each uncertainty component, uncertainty is synthesized to obtain the synthesized standard uncertainty of the online evaluation algorithm for the energy meter error.

[0047] The expanded uncertainty is obtained based on the combined standard uncertainty.

[0048] In a second aspect, the present invention provides an electronic device comprising a memory and a processor, wherein the memory stores a computer program executed by the processor, the computer program, when executed by the processor, causes a device equipped with the processor to perform an uncertainty assessment method for an online evaluation algorithm for electricity meter errors as described in any of the first aspects above.

[0049] A third aspect of the present invention provides a storage medium storing a computer program that runs on a computer and, when running, causes the computer to perform an uncertainty assessment method for an online energy meter error evaluation algorithm as described in any of the first aspects above.

[0050] Compared with the prior art, the beneficial effects of the present invention are: under the premise of ensuring that the online evaluation algorithm for electricity meter error has linear characteristics, the experimental space can be reduced to a very small range. This not only effectively solves the problem that traditional analytical methods cannot perform comprehensive calculations, but also overcomes the limitations of Monte Carlo methods in comprehensive analysis across various dimensions, thereby improving computational efficiency and feasibility. Attached Figure Description

[0051] The above and other objects, features, and advantages of the present invention will become more apparent from the more detailed description of the embodiments of the invention in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof.

[0052] Figure 1 This is a flowchart illustrating the uncertainty assessment method of the online evaluation algorithm for electricity meter errors in some embodiments of the present invention;

[0053] Figure 2 This is a topology distribution diagram of a smart energy meter in some embodiments of the present invention;

[0054] Figure 3 This is a schematic diagram of the process for evaluating the zero-point uncertainty of an algorithm using the Monte Carlo method in some embodiments of the present invention;

[0055] Figure 4 This is a flowchart illustrating the process of verifying the linearity of the algorithm through a linear experiment that solves the problem by varying the total table error in some embodiments of the present invention.

[0056] Figure 5 This is a flowchart illustrating the process of verifying the linearity of the algorithm through a linear experiment that solves the variation of table partitioning error in some embodiments of the present invention.

[0057] Figure 6This is a flowchart illustrating the uncertainty assessment method of the online evaluation algorithm for electricity meter errors in other embodiments of the present invention;

[0058] Figure 7 This is a schematic diagram of the structure of an electronic device according to some embodiments of the present invention. Detailed Implementation

[0059] To make the objectives, technical solutions, and advantages of the present invention more apparent, exemplary embodiments according to the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are merely a part of the embodiments of the present invention, and not all of the embodiments of the present invention. It should be understood that the present invention is not limited to the exemplary embodiments described herein. Based on the embodiments of the present invention described herein, all other embodiments obtained by those skilled in the art without inventive effort should fall within the protection scope of the present invention.

[0060] Currently, evaluation methods for online energy meter error evaluation algorithms typically rely on various error metrics, such as MAPE, RMSE, MSE, RSSE, false negative rate, false positive rate, and detection rate. While these error evaluation metrics are widely used in practical applications, they still have certain limitations in algorithm performance evaluation. Traditional error evaluation methods can usually only provide an overall evaluation of the algorithm, failing to deeply analyze the sources of error or provide the algorithm's uncertainty. Furthermore, existing uncertainty evaluation methods, such as the GUM method and the Monte Carlo method, cannot handle the complex relationships between multiple variables in the algorithm and the influence of multiple uncertain factors in practical applications. Therefore, they cannot effectively evaluate the performance of online energy meter error evaluation algorithms. Thus, to more comprehensively evaluate algorithm performance, this invention proposes an uncertainty evaluation method for online energy meter error evaluation algorithms. This method, based on analytical analysis and Monte Carlo experiments, overcomes the limitations of purely analytical and Monte Carlo methods in analyzing algorithm uncertainty, providing a comprehensive analysis of the algorithm's uncertainty. The invention will be further described in detail below with reference to the accompanying drawings.

[0061] To facilitate understanding of this invention, a detailed description of the uncertainty assessment method for an online energy meter error evaluation algorithm disclosed in this embodiment is provided first. The execution entity of the uncertainty assessment method for the online energy meter error evaluation algorithm provided in this embodiment is generally a computer device with certain computing capabilities. This computer device may include, for example, a terminal device, a server, or other processing devices. The terminal device may be a user equipment (UE), a mobile device, a user terminal, etc. In some possible implementations, the uncertainty assessment method for the online energy meter error evaluation algorithm can be implemented by a processor calling computer-readable instructions stored in memory.

[0062] The following is for reference Figure 1 This invention describes an uncertainty assessment method for an online energy meter error evaluation algorithm in some embodiments of the present invention.

[0063] like Figure 1 As shown, the uncertainty assessment method of the online evaluation algorithm for electricity meter error in this invention includes at least the following:

[0064] S110. Verify the linearity of the online evaluation algorithm for the error of the electricity meter to be evaluated.

[0065] The electricity meter in this invention is a smart electricity meter, which can monitor, record and transmit the user's electricity consumption data in real time.

[0066] The online error evaluation algorithm for electricity meters in this invention is a calibration method that uses data analysis algorithms to calculate errors based on the metering data of smart electricity meters. This method processes data from multiple electricity meters in batches in the background, achieving accurate error calculation and analysis, significantly improving detection efficiency. It also possesses real-time error assessment and operational status monitoring capabilities, thereby accurately locating metering errors and quickly identifying equipment faults.

[0067] It should be noted that this invention does not limit the specific evaluation method of the online evaluation algorithm for electricity meter errors.

[0068] S120. For the verified online evaluation algorithm for electricity meter error, analyze the sources of uncertainty in the online evaluation algorithm for electricity meter error, and decompose the uncertainty of the algorithm into multiple uncertainty components.

[0069] S130. For each uncertainty component, the GUM method or Monte Carlo method is used for evaluation to obtain the evaluation results of each uncertainty component.

[0070] The choice between using the GUM method or the Monte Carlo method to evaluate each uncertainty component depends on the complexity of the model, the type of input distribution, and whether error decomposition or sensitivity analysis is required. If the model is linear or approximately linear, and the distribution of the input variables is relatively simple (e.g., normally distributed), the GUM method can be used. For complex, nonlinear, or non-linear models, especially when the distribution of input variables is irregular or complex, the Monte Carlo method is preferred. In this invention, some components (such as the uncertainty of the summary table itself) can be calculated using prior knowledge based on the GUM method, while the inherent uncertainty components of the algorithm itself can be analyzed using the Monte Carlo method.

[0071] S140. Based on the evaluation results of each uncertainty component, the uncertainty of the online evaluation algorithm for electricity meter error is obtained.

[0072] The method of this invention can be applied to online metering systems for smart meters in low-voltage distributed networks. Online metering systems for smart meters are typically low-voltage network systems with a distributed topology oriented towards the meters. Figure 2 As shown, it includes a centralized master table M0 and sub-tables M1, M2, ... M0 located in multiple branches. n .

[0073] The centralized master meter M0 is located at the entrance of the low-voltage network and is responsible for recording the total power consumption of the entire network. It is an important reference point for energy conservation analysis.

[0074] Subtables M1, M2, ... M n Distributed across different branches, each branch records the energy consumption of its subordinate users 1, 2, ..., n. Each user is connected to a renewable energy source, but this invention does not consider the impact of renewable energy sources on the metering error of the energy meter.

[0075] Subtables M1, M2, ... M n Metering data from the centralized master meter M0 is collected by a collector and then transmitted in real time to the system backend via a communication system. The system backend typically includes a data processing terminal and a database, with the data processing terminal connected to the communication system via a concentrator. The communication system can connect to each smart meter using wired or wireless communication methods. Wired communication methods include, but are not limited to, Ethernet communication and serial communication, while wireless communication methods include, but are not limited to, Wi-Fi communication, Bluetooth communication, and mobile communication networks.

[0076] Each user terminal (such as a home or small business) has its electricity consumption recorded by a corresponding sub-meter. The total electricity consumption of all users must satisfy the law of conservation of energy.

[0077]

[0078] in, It represents the actual electricity consumption of the total meter during the i-th measurement period; ω is the actual electricity consumption of the j-th submeter in the area during the i-th measurement period; line loss ω represents the energy loss caused by the transmission line during the power transmission process in the i-th measurement period; fixed loss This refers to the loss caused by a fixed load that cannot be measured by the measuring device during the i-th measurement period.

[0079] This invention provides an uncertainty assessment method for online energy meter error evaluation algorithms, systematically evaluating the uncertainty of these algorithms. While maintaining linearity, the experimental space can be minimized, thereby improving computational efficiency and feasibility. Ultimately, it achieves uncertainty assessment for a class of online energy meter error evaluation algorithms. This method not only effectively solves the problem of traditional analytical methods being unable to perform comprehensive calculations but also overcomes the limitations of Monte Carlo methods in multi-dimensional comprehensive analysis, providing a scientifically reliable solution for uncertainty assessment of online energy meter error evaluation algorithms. Furthermore, this method is applicable to smart energy meter systems in low-voltage distributed networks, exhibiting good scalability and adaptability to meet the needs of different application scenarios.

[0080] In some embodiments of the present invention, the multiple uncertainty components include the zero-point uncertainty component of the algorithm, the uncertainty component of the summary table, and the uncertainty component of the line loss calculation.

[0081] By analyzing the sources of uncertainty, the uncertainty of an algorithm can be decomposed into multiple uncertainty components. Figure 2 The sources of algorithm uncertainty in the online metering system of the smart energy meter shown are illustrated in Table 1 below:

[0082] Table 1

[0083] Sources of uncertainty Remark Summary table accuracy Main items Current transformer Main items Line loss estimation Main items The power resolution of the main meter Main items The algorithm itself Main items Energy resolution of each sub-meter Secondary items Average error replaces point error Secondary items Clock misalignment Secondary items Impacts of environmental change Secondary items

[0084] Based on existing experience and relevant literature, the uncertainty of the online evaluation algorithm for electricity meter errors in this embodiment can be divided into the uncertainty of the total meter (including current transformers), the uncertainty of line loss calculation, and the uncertainty of the algorithm's zero point. The zero-point uncertainty is the uncertainty arising from the algorithm's inherent properties when the errors of all sub-meters are zero.

[0085] The uncertainty of the total meter (including current transformers) and the uncertainty of line loss calculation can be obtained through prior knowledge. For the zero-point uncertainty component, the Monte Carlo method can be used for evaluation to obtain the zero-point uncertainty of the algorithm.

[0086] Next, refer to Figure 3 This describes the specific steps of evaluating the zero-point uncertainty of an algorithm using the Monte Carlo method in some embodiments of the present invention.

[0087] Before introducing the evaluation steps, we will first explain the zero-error dataset used in the evaluation process.

[0088] The zero-error dataset is constructed based on the topology of the low-voltage distribution area. In this invention, it can be used in the evaluation process of the Monte Carlo method, and also in the verification process of the linearity of the online evaluation algorithm for the error of the electricity meter to be evaluated.

[0089] In some embodiments of the present invention, the method for constructing a zero-error dataset includes:

[0090] Collect metering datasets from the online metering system of smart energy meters in low-voltage distribution areas;

[0091] Based on the metering dataset, an initial zero-error dataset is generated through a simulation data platform, meaning that the error of all meters is zero.

[0092] The following is a specific example of a metering dataset. In this example, the metering dataset comes from a residential user distribution area in a province or city in China. This distribution area contains 160 independent sub-meters and is equipped with a master meter with a magnification factor of 200. All meters have a resolution of 0.01 kWh, and data is collected every 15 minutes, with 96 data points collected daily. The data collection period spans 13 months, covering the complete electricity consumption situation of this low-voltage distribution area.

[0093] Preferably, in order to simulate real transformer area data, the data in the zero-error dataset is reduced to simulate the impact of quantization error on the algorithm.

[0094] like Figure 3 As shown, in some embodiments of the present invention, the Monte Carlo method is used for evaluation, including the following steps S310-S340.

[0095] Step S310: Set a third random noise and load it into the main meter's energy data and the sub-meter's energy data for each sub-meter.

[0096] Specifically, matrices X and Y can be constructed. Matrix X represents a matrix composed of the energy values ​​of each sub-meter, with each column of X representing the energy value of the same sub-meter at different time periods, and each row representing the energy value of different sub-meters at the same time period. Y represents the energy value of the master meter at different time periods.

[0097] Set random noise, for example, random noise with a standard deviation of 0.2%, and add random noise with a standard deviation of 0.2% to each element of matrices X and Y.

[0098] Step S320: The calculation error of each sub-meter is obtained by solving the online evaluation algorithm for the electricity meter error.

[0099] A system of linear equations is constructed based on the law of conservation of energy, and solved using an online evaluation algorithm for electricity meter errors, which yields the calculation error for each meter.

[0100] Step S330, repeat steps S310 and S320 to obtain multiple calculation errors for each sub-table, and calculate the mean and standard deviation of the multiple calculation errors for each sub-table.

[0101] Specifically, 100 experiments can be conducted according to the zero-point uncertainty experimental procedure, and the average value and standard deviation of the solution obtained by the algorithm each time can be calculated.

[0102] Step S340: Calculate the uncertainty of each sub-table according to the following formula, perform statistical analysis on the uncertainty of each sub-table, and obtain the zero-point uncertainty of the algorithm:

[0103]

[0104] Among them, u a The values ​​represent the uncertainty of the sub-tables. 'a' represents the mean, indicating the center of the experiment, i.e., the average level of the algorithm's output under noise interference. 'b' represents the standard deviation, indicating the dispersion of the data distribution and reflecting the volatility of the solution after noise is introduced. A larger standard deviation means the algorithm is more sensitive to noise, and the zero-point uncertainty increases accordingly. 'max' represents taking the larger of |a+2b| and |a-2b|.

[0105] Furthermore, starting with the uncertainty of the algorithm's zero point, the zero-error dataset is applied to gradually adjust the impact of random noise on the zero-error dataset, simulating all factors that may affect the solution accuracy, such as the power resolution of the main table, the algorithm itself, the power resolution of each sub-table, the average error replacing the error at each point, clock misalignment, and the impact of environmental changes.

[0106] By reducing the error space to a subspace consisting of zeros, where both the total table and sub-table errors are zero, the limitation of Monte Carlo methods requiring extensive simulation experiments in multidimensional space can be overcome. Subsequently, the linearity of the algorithm can be used to extend this subspace to any other error space.

[0107] In some embodiments of the present invention, the linearity of the online evaluation algorithm for the electricity meter error is verified experimentally. Specifically, based on the zero-error dataset, the linearity of the online evaluation algorithm for the electricity meter error can be verified through linear experiments on the solution of the total meter error and linear experiments on the solution of the solution of the individual meter errors.

[0108] In the linear experiment of the solution changing with the total meter error, the linearity of the online evaluation algorithm for the energy meter error under evaluation is verified based on the zero error dataset. By adjusting the total meter error, the change of the algorithm solution is observed to see whether it has a linear response, thus ensuring that it meets the precondition of linear superposition.

[0109] Next, refer to Figure 4 This describes the linearity of the algorithm in some embodiments of the present invention, verified by linear experiments on the solution of the total table error.

[0110] like Figure 4As shown, the linearity of the algorithm is verified through a linear experiment that solves the problem by varying the total table error. The steps include:

[0111] Step S410: Randomly set the first increment factor to adjust the total meter power data.

[0112] Specifically, the first increment factor λ can be randomly set to adjust Y, resulting in:

[0113] Y ′ =Y×(1+λ)

[0114] The error in the summary table was modified by adjusting Y using an increment factor.

[0115] In this experiment, a first increment factor λ, randomly generated from a normal distribution with a mean of 0 and a standard deviation of 2%, is applied to the summary table data to modify the error of the summary table. The summary table data here is the data obtained after reducing the fractions of the zero-error dataset generated by the simulation data platform.

[0116] The master data generated by the simulation data platform learns the electricity consumption characteristics and habits of each user in the real distribution area data. The resulting dataset construction reflects the characteristics of the real distribution area data while ensuring controllable errors in the distribution area's electricity meter data. During verification, the linearity of the online evaluation algorithm for the evaluated electricity meter errors is verified based on the zero-error dataset, thus ensuring that it meets the prerequisite of linear superposition.

[0117] Step S420: Set the first random noise and load the sub-meter energy data and the adjusted total meter energy data into each sub-meter.

[0118] Specifically, random noise can be set and loaded onto each element of matrices X and Y′.

[0119] For example, random noise can be random noise with a standard deviation of 0.2%.

[0120] Step S430: Solve the error online evaluation algorithm of the electricity meter, and obtain the first deviation value by subtracting the theoretical value from the actual value obtained by the algorithm.

[0121] Step S440: Repeat steps S410-S430 above, and calculate the statistical results of the first deviation value;

[0122] Step S450: Based on the statistical results of the first deviation value, verify the linearity of the change in total meter error solved by the online evaluation algorithm for electricity meter error.

[0123] During the verification process, it is necessary to analyze the linearity of the algorithm's solution as the total table error changes.

[0124] When analyzing its linearity, the average value of the solution obtained by each energy meter can be compared with the value of the corresponding element (theoretical value) of the set first increment factor λ. The smaller the deviation, the better the linearity.

[0125] Because substituting Y′ into the equation yields a new solution, which in turn yields a new round of electricity meter error values.

[0126] Y′=Xθ′

[0127] in:

[0128]

[0129] Subtracting the original system of equations from the new system of equations yields:

[0130] λ×Y=X(θ′-θ)

[0131] Easy to obtain:

[0132]

[0133] Since (1-σ1) is approximately 1, we can obtain:

[0134]

[0135] Where σ1′ represents the error value of the initial solution, σ1 represents the error value of the algorithm after adding the increment factor, and n represents the number of sub-meters of the electricity meter.

[0136] During the data preparation phase, the errors of each sub-table are set to zero. When a linear relationship exists between Y and the solution, incremental adjustments to Y should reflect a linear response in the solution of the equation and be reflected in the error obtained from the solution. That is, the difference between the error of each sub-table obtained after incremental adjustment and the initial solution (the solution obtained without incremental adjustment) is λ. This difference reflects the deviation from the initial solution, thereby verifying whether the system maintains its linear response characteristics.

[0137] Specifically, under the condition of repeatability measurement, N experiments can be conducted, with N ranging from 10 to 30 (e.g., N = 20). The deviation between the theoretical and actual values ​​of each experiment is obtained, and then the mean and variance of the deviation values ​​corresponding to each table are calculated.

[0138] In a specific example, statistical analysis showed that 97% of the deviation values ​​had an average value within ±0.5%, indicating that the change in the solution after incrementally adjusting the total table error was consistent with the theoretical prediction. Experimental analysis showed that when the total table error changed, the error value calculated by the algorithm exhibited a linear response characteristic with the change in the incremental factor, and the amount of error change was basically consistent with the theoretical prediction. Therefore, the solution obtained by the algorithm had a linear relationship with the change in the total table error.

[0139] The linear experiment of the solution changing with the table partitioning error is conducted by adjusting the table partitioning error and observing whether the change of the algorithm solution is linear.

[0140] Next, refer to Figure 5 This describes the linearity of the algorithm in some embodiments of the present invention, verified by linear experiments on the solution of the variation of table subtraction error.

[0141] like Figure 5 As shown, the linearity of the algorithm is verified through a linear experiment that solves the problem by varying the table partitioning error. The steps include:

[0142] Step S510: Randomly set a second increment factor to adjust the sub-meter power data of each sub-meter.

[0143] The table data here is obtained by reducing the fractions in the zero-error dataset generated by the simulation data platform.

[0144] Specifically, a second increment factor Δ is randomly set to adjust X.

[0145] In this experiment, the second increment factor [Δ1, Δ2, ..., Δ] was randomly generated from a normal distribution with a mean of 0 and a standard deviation of 0.5%. n This is applied to the partitioned data, thereby changing the partitioning error.

[0146] Multiply each column of X by the increment factor (1+Δ), leaving the other parameters unchanged, to obtain X′:

[0147]

[0148] To verify the linearity of the solution as X changes, an increment vector (1+Δ) is set. By multiplying matrix X by this increment vector, each column vector of matrix X can be adjusted. Each column is multiplied by a different increment factor, and the response of the solution is observed. If the column changes of X are linearly related to the solution, then this incremental adjustment should be directly reflected in the solution of the linear equation. In each adjustment, if the difference between the new solution after the column transformation of matrix X and the initial solution corresponds to the corresponding element in the increment vector (1+Δ), it indicates that the system maintains a linear response to the column adjustments of X. This method ensures that the system retains its linear characteristics when the input changes and does not introduce nonlinear effects due to column adjustments.

[0149] Step S520: Set a second random noise and load it into the main meter's energy data and the adjusted sub-meter's energy data, respectively.

[0150] For example, random noise can be random noise with a standard deviation of 0.2%. This random noise with a standard deviation of 0.2% is loaded onto each element of matrices X' and Y.

[0151] Step S530: Solve the online evaluation algorithm for the energy meter error, and subtract the theoretical value from the actual value obtained by the algorithm to obtain the second deviation value of each sub-meter.

[0152] A series of solutions are obtained by applying the algorithm; a theoretical value for the algorithm error is set, and the difference between the theoretical value and the actual value obtained by the algorithm is calculated to obtain the second deviation value of each sub-table.

[0153] Step S540: Repeat steps S510-S530 above, and calculate the statistical result of the second deviation value.

[0154] Step S550: Based on the statistical results of the second deviation value, verify the linear characteristics of the energy meter error online evaluation algorithm as a function of sub-meter error.

[0155] By analyzing the linearity of the solution obtained by the algorithm as the sub-meter error changes, the linearity of the solution obtained by the online evaluation algorithm for electricity meter error as the sub-meter error changes is verified.

[0156] When analyzing its linearity, the average value of the solution obtained by each energy meter can be compared with the value of the corresponding element of the set incremental factor Δ (theoretical value). The smaller the deviation, the better the linearity.

[0157] Since the other variables remain unchanged, the original system of equations becomes:

[0158] Y=X′θ′

[0159] Subtracting the new system of equations from the original system of equations and simplifying, we get:

[0160]

[0161] Since the error value is very small, almost close to 0, the right side of the equation is approximately -Δ, so we can obtain:

[0162]

[0163] in:

[0164]

[0165] The above method can be used to verify whether the system maintains a linear response to changes in X. According to the characteristics of linear systems, if the solution obtained by the algorithm exhibits a linear response to adjustments in each column of X, then changes in the selected columns should be directly reflected in the solution. This direct linear response indicates that the system maintains linearity to adjustments in each column of the coefficient matrix, and no nonlinear effects are introduced due to column adjustments.

[0166] Specifically, under the condition of repeatability measurement, N experiments can be conducted, for example, N=20, to obtain the deviation between the theoretical value and the actual value in each experiment, and then the mean and variance of the deviation value corresponding to each table can be calculated.

[0167] In a specific example, statistical analysis of the results showed that the average value of 94% of the deviation values ​​was within the range of [-0.5%, 0.5%]. This indicates that after incremental adjustments to each sub-table, the algorithm's calculation results were consistent with the theoretical results, with no obvious systematic deviation. Therefore, the solution obtained by the algorithm is linearly related to the change in sub-table error.

[0168] In summary, the analysis of the two experimental results shows that the algorithm satisfies linearity verification and possesses good linearity properties. Therefore, the error space can be reduced to a subspace consisting of zeros, where both the total table and sub-table errors are zero, thus solving for the zero-point uncertainty of the algorithm. This effectively overcomes the limitation of the Monte Carlo method, which requires extensive simulation experiments in a multi-dimensional space.

[0169] Next, refer to Figure 6 This invention describes an uncertainty assessment method for an online energy meter error evaluation algorithm in some embodiments of the present invention.

[0170] S610. Verify the linearity of the online evaluation algorithm for the error of the electricity meter to be evaluated.

[0171] S620. For the verified online evaluation algorithm for electricity meter error, analyze the sources of uncertainty in the online evaluation algorithm for electricity meter error, and decompose the uncertainty of the algorithm into multiple uncertainty components.

[0172] S630. For each uncertainty component, the GUM method or Monte Carlo method is used for evaluation to obtain the evaluation results of each uncertainty component.

[0173] The uncertainties in the summary table (including current transformers) and line loss calculation can be obtained through prior knowledge. An experiment was designed to solve the zero-point uncertainty of the algorithm, and the solution results are shown in Table 2.

[0174] Table 2

[0175] Uncertainty components Standard uncertainty (%) Sensitivity coefficient Summary meter (including current transformer) <![CDATA[u1]]> 1 Uncertainty in line loss calculation <![CDATA[u2]]> 1 Zero-point uncertainty <![CDATA[u3]]> 1

[0176] S640. Based on the evaluation results of each uncertainty component, perform uncertainty synthesis to obtain the synthesized standard uncertainty of the online evaluation algorithm for the energy meter error.

[0177]

[0178] Among them, u c (y) represents the combined standard uncertainty.

[0179] S650. Based on the combined standard uncertainty, the expanded uncertainty is obtained.

[0180] The expanded uncertainty of the algorithm is:

[0181] U = u c (y)×k (k=2)

[0182] Where U is the expanded uncertainty and k is the coverage factor, which is used to expand the combined standard uncertainty into an expanded uncertainty to reflect the reliability of the measurement results.

[0183] The method described in this embodiment can be applied to electricity metering devices to systematically evaluate the uncertainty of online evaluation algorithms for electricity meter errors, assessing the accuracy and reliability of the algorithms. It offers the following advantages:

[0184] 1. Fill the gap in online evaluation algorithms for electricity meter errors.

[0185] Currently, research on uncertainty assessment of online evaluation algorithms for electricity meter errors, both domestically and internationally, is still in the exploratory stage, and a systematic methodology has not yet been established. This method fills the research gap in the field of uncertainty assessment for online evaluation algorithms for electricity meter errors, providing a systematic and feasible uncertainty assessment method, and offering scientific support for the accuracy and reliability of online evaluation of electricity meter errors.

[0186] 2. Improve the scientific rigor of algorithm evaluation.

[0187] Through linear verification tests based on a zero-error standard dataset, the evaluated online error estimation algorithm for electricity meters is ensured to have linear characteristics. This allows for the use of linear superposition in uncertainty assessment, improving the rationality and operability of the algorithm's uncertainty calculation.

[0188] 3. Sources of uncertainty in system analysis

[0189] This invention provides an in-depth analysis of the influencing factors and uncertainty sources of online evaluation algorithms for electricity meter errors. It decomposes the algorithm's uncertainty into multiple uncertainty components. Different uncertainty components are evaluated using either the GUM method or the Monte Carlo method, thereby making the evaluation more targeted and accurate.

[0190] 4. Overcoming the limitations of traditional methods

[0191] For the uncertainty assessment problem of online evaluation algorithms for electricity meter errors, traditional analytical methods are difficult to perform comprehensive calculations. Although applicable to complex uncertainty analysis, they are time-consuming and costly. The algorithm uncertainty assessment method of this invention makes the algorithm feasible in a very small experimental space by performing linear analysis, thereby overcoming the shortcomings of existing methods and achieving a more comprehensive and accurate uncertainty assessment.

[0192] 5. Improve the reliability of online monitoring of electricity meters.

[0193] By systematically evaluating the uncertainty of the online evaluation algorithm for electricity meter errors, this invention improves the reliability of the online estimation algorithm for electricity meter errors, which helps the power system achieve more efficient online monitoring and calibration, reduces maintenance costs, and improves the accuracy of electricity metering data.

[0194] The method of this invention can effectively improve the online monitoring and calibration level of power metering devices, and ensure the accuracy and stability of power metering in the power system.

[0195] This application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program that is executed by the processor. When the computer program is executed by the processor, it causes the device equipped with the processor to perform the uncertainty assessment method of the online evaluation algorithm for electricity meter error as described in any of the above embodiments.

[0196] Next, refer to Figure 7 An example electronic device 700 for implementing the methods of embodiments of the present invention is described below.

[0197] like Figure 7 As shown, the electronic device 700 includes a processor 710, a memory 720, and a communication interface 730. The processor 710, memory 720, and communication interface 730 can be interconnected and communicate via a communication bus 740 and / or other forms of connection mechanisms (not shown).

[0198] It should be noted that Figure 7 The components and structure of the electronic device 700 shown are merely exemplary and not limiting; the electronic device may also have other components and structures as needed.

[0199] Optionally, the communication interface 730 may also include a transmitter and / or a receiver.

[0200] The processor 710 may be a microcontroller unit (MCU), a central processing unit (CPU), a digital signal processor (DSP), a microcontroller, an embedded device, or other processing units with data processing and / or instruction execution capabilities.

[0201] The memory 720 can be various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM), cache memory, synchronous dynamic random access memory (SDRAM), etc. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may also be stored on the computer-readable storage medium, and the memory 720 can execute the program instructions to implement the uncertainty evaluation method in the embodiments of the present invention described below.

[0202] This application also provides a storage medium storing a computer program that runs on a computer. When the computer program runs, it causes the computer to execute the uncertainty assessment method of the online evaluation algorithm for electricity meter error as described in any of the above embodiments.

[0203] Finally, it should be noted that the above technical solution is only one embodiment of the present invention. For those skilled in the art, based on the application methods and principles disclosed in the present invention, it is easy to make various types of improvements or modifications, and not limited to the methods described in the above specific embodiments of the present invention. Therefore, the methods described above are only preferred and have no limiting significance.

Claims

1. An uncertainty assessment method for an online evaluation algorithm for electricity meter errors, characterized in that: The uncertainty assessment method includes: The linearity of the online evaluation algorithm for the error of the electricity meter to be evaluated was verified; For the verified online evaluation algorithm for electricity meter error, the sources of uncertainty in the algorithm are analyzed, and the uncertainty of the algorithm is decomposed into multiple uncertainty components. For each uncertainty component, the GUM method or Monte Carlo method is used for evaluation to obtain the evaluation results of each uncertainty component; Based on the evaluation results of each uncertainty component, the uncertainty of the online evaluation algorithm for electricity meter error is obtained.

2. The uncertainty assessment method for the online evaluation algorithm of electricity meter error according to claim 1, characterized in that: The linearity of the online evaluation algorithm for the electricity meter error to be evaluated was verified, including: Based on the topology of the online metering system for electricity meters in the low-voltage distribution area, an initial zero-error dataset is constructed, which includes the total meter electricity data and the sub-meter electricity data of each sub-meter. Based on the zero-error dataset, the linearity of the online energy meter error evaluation algorithm is verified through linear experiments on the solution varying with the total meter error and linear experiments on the solution varying with the sub-meter error.

3. The uncertainty assessment method for the online evaluation algorithm of electricity meter error according to claim 2, characterized in that: The linearity of the online energy meter error evaluation algorithm is verified by solving a linear experiment that varies with the total meter error. The steps include: Step S410: Randomly set a first increment factor to adjust the total meter power data; Step S420: Set the first random noise and load the sub-meter energy data and the adjusted total meter energy data into each sub-meter respectively; Step S430: Solve the error online evaluation algorithm of the electricity meter, and obtain the first deviation value by subtracting the theoretical value from the actual value solved by the algorithm; Step S440: Repeat steps S410-S430, and calculate the statistical results of the first deviation value; Step S450: Based on the statistical results of the first deviation value, verify the linearity of the change in total meter error solved by the online evaluation algorithm for electricity meter error.

4. The uncertainty assessment method for the online evaluation algorithm of electricity meter error according to claim 2, characterized in that: The linearity of the online energy meter error evaluation algorithm is verified through a linear experiment that solves the problem of how the meter error changes, including: Step S510: Randomly set the second increment factor to adjust the sub-meter power data of each sub-meter; Step S520: Set a second random noise and load it into the total meter energy data and the adjusted sub-meter energy data respectively; Step S530: Solve the error online evaluation algorithm of the electricity meter, and subtract the theoretical value of the algorithm from the actual value obtained by the algorithm to obtain the second deviation value of each sub-meter; Step S540: Repeat steps S510-S530, and calculate the statistical result of the second deviation value; Step S550: Based on the statistical results of the second deviation value, verify the linear characteristics of the energy meter error online evaluation algorithm as a function of sub-meter error.

5. The uncertainty assessment method for the online evaluation algorithm of electricity meter error according to claim 2, characterized in that: Multiple uncertainty components include the zero-point uncertainty component of the algorithm, the uncertainty component of the summary table, and the uncertainty component of the line loss calculation.

6. The uncertainty assessment method for the online evaluation algorithm of electricity meter error according to claim 5, characterized in that: The zero-point uncertainty component is evaluated using the Monte Carlo method to obtain the zero-point uncertainty of the algorithm.

7. The uncertainty assessment method for the online evaluation algorithm of electricity meter error according to claim 6, characterized in that: The steps involved in the evaluation using the Monte Carlo method include: Step S310: Set the third random noise and load it into the main meter's energy data and the sub-meter's energy data for each sub-meter. Step S320: Solve the calculation error of each sub-meter using the online evaluation algorithm for electricity meter error; Step S330, repeat steps S310 and S320 to obtain multiple calculation errors for each sub-table, and calculate the mean and standard deviation of the multiple calculation errors for each sub-table. Step S340: Calculate the uncertainty of each sub-table using the following formula, perform statistical analysis on the uncertainty of each sub-table, and obtain the zero-point uncertainty of the algorithm: Among them, u a The expression represents the uncertainty of the sub-table, a represents the mean, b represents the standard deviation, and max represents taking the larger of |a+2b| and |a-2b|.

8. The uncertainty assessment method for the online evaluation algorithm of electricity meter error according to claim 1, characterized in that: Based on the evaluation results of each uncertainty component, the uncertainty of the online evaluation algorithm for electricity meter error is obtained, including: Based on the evaluation results of each uncertainty component, uncertainty is synthesized to obtain the synthesized standard uncertainty of the online evaluation algorithm for the energy meter error. The expanded uncertainty is obtained based on the combined standard uncertainty.

9. An electronic device, characterized in that, The electronic device includes a memory and a processor, the memory storing a computer program executed by the processor, the computer program, when executed by the processor, causing the device equipped with the processor to perform the uncertainty assessment method of the online evaluation algorithm for electricity meter errors as described in any one of claims 1-8.

10. A storage medium, characterized in that, The storage medium stores a computer program that runs on a computer. When the computer program runs, it causes the computer to perform the uncertainty assessment method of the online evaluation algorithm for electricity meter errors as described in any one of claims 1-8.

Images