A method and device for analyzing metering errors of an electric energy meter

Abstract

The present application relates to the technical field of electric energy metering, and in particular to an electric energy metering error analysis method and device. The method comprises: obtaining characteristic data of a current electric energy meter; wherein the characteristic data comprises power grid data, load data, environmental data and operation data; determining a multi-dimensional feature matrix based on the power grid data, load data and environmental data; inputting the multi-dimensional feature matrix into a preset neural network model to obtain a first metering error; determining a second metering error based on the operation data; and determining a third metering error based on the first metering error and the second metering error. In this way, the present application can improve the accuracy of calculating the electric energy metering error.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electric energy metering, and in particular to an electric energy metering error analysis method and device. BACKGROUND

[0002] As an electric power system "trade settlement fair, energy settlement" electric power equipment, the measurement accuracy of electric energy metering is directly related to the economic interests of power supply enterprises and users, and also affects the accuracy of key businesses such as power grid line loss analysis and new energy consumption monitoring. With the transformation of the electric power system to "high penetration of new energy, diversified load, and intelligent power grid", the traditional electric energy metering error control technology has been difficult to adapt to the complex and changing field operation environment.

[0003] Currently, electric energy metering equipment has evolved from early induction type mechanical electric energy meters and traditional electronic electric energy meters to intelligent electric energy meters integrating metering, communication, and edge computing functions. However, the measurement error of intelligent electric energy meters is affected by multi-dimensional factor coupling, and the error data obtained by the existing standard table comparison method can only reflect the basic metering capability of the electric meter under ideal working conditions, resulting in inaccurate calculation of the metering of the electric meter.

[0004] Therefore, the present application provides an electric energy metering error analysis method and device to solve the above technical problems. SUMMARY

[0005] The present application describes an electric energy metering error analysis method and device, which can improve the accuracy of calculating the electric energy metering error.

[0006] According to a first aspect, the present application provides an electric energy metering error analysis method, which comprises:

[0007] obtaining feature data of a current electric energy meter; wherein the feature data includes power grid data, load data, environmental data, and operation data;

[0008] determining a multi-dimensional feature matrix based on the power grid data, the load data, and the environmental data;

[0009] inputting the multi-dimensional feature matrix into a preset neural network model to obtain a first metering error;

[0010] determining a second metering error based on the operation data;

[0011] determining a third metering error based on the first metering error and the second metering error.

[0012] According to a second aspect, the present application provides an electric energy metering error analysis device, comprising:

[0013] An acquisition unit is configured to acquire feature data of a current electric energy meter; wherein the feature data comprises power grid data, load data, environment data, and operation data;

[0014] A first data processing unit is configured to determine a multi-dimensional feature matrix based on the power grid data, the load data, and the environment data;

[0015] A second data processing unit is configured to input the multi-dimensional feature matrix into a preset neural network model to obtain a first measurement error;

[0016] A third data processing unit is configured to determine a second measurement error based on the operation data;

[0017] A fourth data processing unit is configured to determine a third measurement error based on the first measurement error and the second measurement error.

[0018] In a third aspect, an electronic device is provided, including a memory and a processor, the memory stores a computer program, and the processor executes the computer program to implement the method in any of the embodiments of the present specification.

[0019] In a fourth aspect, a computer readable storage medium is provided, which stores a computer program, and the computer program, when executed in a computer, causes the computer to execute the method in any of the embodiments of the present specification.

[0020] According to the electric energy meter measurement error analysis method and device, the feature data of the current electric energy meter is acquired; wherein the feature data comprises power grid data, load data, environment data, and operation data; the multi-dimensional feature matrix is determined based on the power grid data, the load data, and the environment data; the multi-dimensional feature matrix is input into the preset neural network model to obtain the first measurement error; the preset neural network model uses the strong fitting ability of the model for the nonlinear relationship to output the first measurement error reflecting the comprehensive influence of the complex field environment (such as new energy access and load fluctuation); the limitation that the traditional standard table comparison method is only suitable for ideal working conditions is solved; at the same time, the second measurement error is independently calculated based on the operation data; the second measurement error is used to represent the measurement deviation caused by the change of the physical characteristics of the device itself (such as temperature drift and device aging); the third measurement error and the final calculation result are determined according to the first measurement error and the second measurement error. In this way, the accuracy of calculating the electric energy meter measurement error can be improved. BRIEF DESCRIPTION OF DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the accompanying drawings needed to be used in the embodiments or prior art description will be briefly introduced as follows. Obviously, the accompanying drawings in the following description are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative effort based on these drawings.

[0022] Figure 1 A flowchart of an electric energy metering error analysis method according to one embodiment is shown;

[0023] Figure 2 A schematic block diagram of an electric energy metering error analysis device according to one embodiment is shown. DETAILED DESCRIPTION

[0024] The schemes provided by the present application will be described below with reference to the accompanying drawings.

[0025] Figure 1 A flowchart of an electric energy metering error analysis method according to one embodiment is shown. It can be understood that the method can be executed by any device, equipment, platform, device cluster with computing and processing capabilities. As shown, the method comprises: Figure 1

[0026] Step 100, obtaining feature data of a current electric energy meter; wherein the feature data comprises power grid data, load data, environmental data and operation data;

[0027] Step 102, determining a multi-dimensional feature matrix based on the power grid data, the load data and the environmental data;

[0028] Step 104, inputting the multi-dimensional feature matrix into a preset neural network model to obtain a first metering error;

[0029] Step 106, determining a second metering error based on the operation data;

[0030] Step 108, determining a third metering error based on the first metering error and the second metering error.

[0031] ​In the embodiment, the characteristic data of the current electric energy meter is acquired; wherein the characteristic data comprises power grid data, load data, environment data and operation data; the multi-dimensional characteristic matrix is determined based on the power grid data, the load data and the environment data; the multi-dimensional characteristic matrix is input into a preset neural network model to obtain a first measurement error; the preset neural network model utilizes the strong fitting capability of the model for a nonlinear relationship to output the first measurement error which can reflect the comprehensive influence of complex field environments (such as new energy access, load fluctuation, etc.), and solves the limitation that the traditional standard table comparison method is only applicable to ideal working conditions; meanwhile, a second measurement error is independently calculated based on the operation data; the second measurement error is used to represent the measurement deviation caused by the change of the physical characteristics of the equipment itself (such as temperature drift, device aging), and then the third measurement error is determined according to the first measurement error and the second measurement error, which is the final calculation result. In this way, the accuracy of calculating the measurement error of the electric energy meter can be improved.

[0032] In the embodiment, the multi-dimensional characteristic matrix is input into a preset neural network model to obtain a first measurement error, and the fitting capability of the neural network for a nonlinear relationship is fully utilized. The preset model needs to be trained by a large amount of historical data (such as the corresponding relationship between the characteristic data under different working conditions and the actual error), and can learn the error law under the coupling action of multiple factors such as power grid fluctuation, load change and environmental interference. Therefore, the actual measurement deviation of the electric energy meter in the complex field environment can be effectively reflected.

[0033] In an embodiment of the present application, the power grid data comprises voltage, voltage deviation rate, frequency and harmonic content; the load data comprises current, load rate and power factor; the environment data comprises environmental temperature and relative humidity; and the operation data comprises the actual value of the sampling resistor, the actual transformation ratio of the current transformer, the actual value of the voltage dividing resistor, the quantization error of the ADC chip and the actual power grid voltage.

[0034] In the embodiment, the specific parameter selection of the power grid data, the load data, the environment data and the operation data is determined by the influence factors of the electric energy meter measurement principle. The voltage deviation rate and the frequency reflect the influence of the stability of the power grid on the measurement reference; the harmonic content is related to the measurement accuracy under nonlinear load; the load rate and the power factor affect the working point of the measurement element; the environmental temperature and humidity directly cause the parameter drift of the components; and the sampling resistor, the current transformer and the ADC chip are the main error sources of the current and voltage measurement link. The selection of these parameters ensures the comprehensiveness of the characteristic data.

[0035] In an embodiment of the present application, the second measurement error is determined based on the operation data, comprising:

[0036] The first error influence factor is determined based on the actual value of the sampling resistor and the actual transformation ratio of the current transformer;

[0037] The second error influence factor is determined based on the actual value of the voltage dividing resistor, the quantization error of the ADC chip and the actual grid voltage.

[0038] The second measurement error is determined based on the first error influence factor and the second error influence factor.

[0039] In the embodiment, first, the first error influence factor is calculated according to the actual value of the sampling resistor and the actual transformation ratio of the current transformer, and then the second error influence factor is determined based on the actual value of the voltage dividing resistor, the quantization error of the ADC chip and the actual grid voltage. Finally, the second measurement error is obtained by integrating the first error influence factor and the second error influence factor, and the second measurement error is used to represent the measurement deviation caused by the change of the physical characteristics of the device itself (such as temperature drift and device aging).

[0040] In an embodiment of the application, the second measurement error is determined by the following formula:

[0041]

[0042] In the formula, is the first error influence factor, is the actual value of the sampling resistor, is the nominal value of the sampling resistor, is the temperature coefficient of the sampling resistor, is the deviation between the actual temperature and the nominal temperature, is the actual transformation ratio of the current transformer, is the nominal transformation ratio of the current transformer, is the preset non-linear error of the current transformer, is the second error influence factor, is the actual value of the voltage dividing resistor, is the nominal value of the voltage dividing resistor, is the temperature coefficient of the voltage dividing resistor, is the quantization error of the ADC chip, is the zero drift error of the ADC chip, is the actual measured grid voltage, is the gain error of the ADC chip, is the second error influence factor, is the second measurement error, is the inherent error.

[0043] In this embodiment, the first error influence factor is used to reflect the error of the current loop, and the second error influence factor is used to reflect the error of the voltage loop. The creativity of this formula lies in both using the square root of the sum of squares to reflect the orthogonality of the current and voltage loop errors, and incorporating inherent errors, which improves the accuracy of calculating the second measurement error. Among them, the temperature coefficient of the sampling resistor (e.g., ±25ppm / ℃, that is, the resistance changes by 0.000025 for every 1℃ change in temperature), the deviation between the actual temperature and the nominal temperature. For example, if the nominal temperature is designed to be 25℃ and the actual temperature is 50℃, then the deviation between the actual temperature and the nominal temperature = 25℃, the zero drift error of the ADC chip (e.g., ±2mV, a fixed deviation that exists regardless of the input voltage), the gain error of the ADC chip (e.g., ±0.05%, the proportional deviation during the input signal amplification process), and the inherent error, that is, the fixed error in the above equation, can be obtained by linear fitting through historical data.

[0044] In one embodiment of the present invention, determining a third measurement error based on a first measurement error and a second measurement error includes:

[0045] The third measurement error is obtained by multiplying the first measurement error by the first weight and the second measurement error by the second weight. The sum of the first weight and the second weight is one. The first weight is used to characterize the confidence level of the first measurement error, and the second weight is used to characterize the confidence level of the second measurement error.

[0046] In this embodiment, the first weight and the second weight are adaptively adjusted based on the confidence levels of the two error calculation methods. When the variance of the first measurement error is small (i.e., the model prediction is more stable), the first weight increases; conversely, if the variance of the second measurement error is smaller (i.e., the hardware parameter calculation is more reliable), the second weight increases. This weighting method based on the statistical characteristics of the data avoids the subjectivity of fixed weights, allowing the third measurement error to preferentially adopt more reliable error sources, thus improving the accuracy of the calculation result (the third measurement error).

[0047] In one embodiment of the present invention, the first weight is determined by the following formula:

[0048]

[0049] In the formula, As the first weight, The first measurement variance, For the first The difference between the model prediction error and the actual error in the first measurement error sample. This is the average of all deviations in the first measurement error sample. The number of samples for the first measurement error. The second measurement variance, The number of samples for the second measurement error. is an average value of all deviations of the second sample, is an average value of all deviations of the second sample.

[0050] In the embodiment, the creativity of the two metrological error variance formulas is to refer to the high-level standard metering value, to calculate the real error of the electric meter to be analyzed first, and then to calculate the deviation based on the real error, so that the variance is no longer a meaningless number, but reflects the deviation fluctuation of the error analysis method from the real situation, meets the professional needs of metrological calibration, and uses completely consistent logic to calculate the variance for the first error of the neural network model and the second error of the hardware formula. The stability of the two types of errors has quantitative comparison indexes, and it can be judged which method is more reliable. The unbiased estimate of "divided by n-1" is used in small sample testing to avoid underestimating the error fluctuation, to fit the actual situation of limited samples on site, and to improve the accuracy of the calculation result (the third metrological error).

[0051] In an embodiment of the present application, before determining the third metrological error based on the first metrological error and the second metrological error, the method further comprises:

[0052] When the absolute value of the difference between the first metrological error and the second metrological error is greater than the first preset value, the step of "obtaining the feature data of the current electric energy meter" is re-executed until the number of cycles reaches the second preset value.

[0053] In the embodiment, when the difference between the first metrological error and the second metrological error exceeds the first preset value, the data is reacquired, which is an abnormality checking procedure. The purpose of the procedure is to exclude abnormal data caused by accidental factors (such as transient power grid fluctuations, temporary sensor failures, etc.), and to ensure the accuracy of the input feature data. The second preset value is set to limit the number of cycles, so as to avoid infinite loop, and through limited retries, reliable data is obtained as much as possible, and finally the accuracy and timeliness of calculating the third metrological error are improved.

[0054] In an embodiment of the present application, when the third metrological error is greater than a third preset value, a correction coefficient of a first weight is obtained, the correction coefficient is multiplied by the first weight to obtain a corrected weight, and the first weight in the step of "multiplying the first metrological error by the first weight and the second metrological error by the second weight to calculate the third metrological error" is replaced by the corrected weight to obtain a fourth metrological error.

[0055] In the embodiment, when the third metrology error is greater than the third preset value, it means that the current fusion result has exceeded the acceptable error range, and the weight correction mechanism needs to be started to optimize the result. Specifically, a correction coefficient of the first weight is obtained, which can dynamically reflect the actual credibility of the first metrology error in the current scene; then the correction coefficient is multiplied by the original first weight to obtain a corrected weight, so as to adjust the proportion of the first metrology error in the fusion calculation; finally, the corrected weight is used to replace the original first weight, and the calculation of "the first metrology error multiplied by the corrected weight plus the second metrology error multiplied by the second weight" is re-executed to obtain a fourth metrology error. In this way, the problem of large error caused by unreasonable initial weight in extreme working conditions is effectively avoided, and the reliability of metrology error calculation is further improved.

[0056] In an embodiment of the application, the corrected weight is determined by the following formula:

[0057]

[0058] In the formula, is the corrected weight, is a weight factor of the harmonic scene, is a harmonic cumulative integral term, is a weight factor of the load scene, is a load rate at the current moment, is a weight factor of the temperature scene, is a temperature offset cumulative integral term, is a harmonic content at the current moment, is a harmonic memory constant, is an ambient temperature, is a historical moment, is a current moment, is an exponential decay factor, is a temperature decay factor, is a nonlinear coefficient, is a limiting function.

[0059] In the present embodiment, the creativity of the above formula is as follows: firstly, multi-scene dynamic fusion breaks through the limitation of the single factor static weight in the prior art, and integrates three types of electric energy meter working condition parameters, i.e., harmonics, load rate and temperature, so that the weight distribution is no longer single and can respond to the complex scene of multi-factor coupling of the power grid; secondly, cumulative effect, the harmonic term reflects the characteristics of "recent high harmonic influence is stronger and long-term influence decays" through integration and exponential decay factor, and the temperature term accurately captures the physical law of "slow change of temperature drift and hysteresis memory" through integration, nonlinear coefficient and decay factor, solving the problem of insufficient description of "long-term cumulative error source" in the traditional method; thirdly, engineering practicability is strengthened, the load rate as a real-time term quickly responds to instantaneous load change, each scene is flexibly adjusted by the weight factor, and the limiting function ensures that the weight is in a reasonable range, avoiding the overflow of the weight in the extreme working condition, which leads to inaccurate error analysis, and finally improving the accuracy of calculating the third measurement error.

[0060] In an embodiment of the present application, when the number of cycles reaches the second preset value, the third measurement error of each cycle is counted, and the third measurement error of each cycle is averaged to obtain a final result.

[0061] In the present embodiment, the traditional electric energy meter error analysis often relies on single calculation, which is easily disturbed by random factors (such as instantaneous power grid fluctuation and temporary device noise), resulting in deviation of the result. When the number of cycles reaches the preset value, the third measurement error obtained multiple times is averaged, which can effectively weaken the influence of random error: multiple cycles can include more rich working condition combinations, and statistical average can offset accidental deviation, the final result is closer to the long-term stable level of the real error, and the accuracy of calculating the third measurement error is finally improved.

[0062] The above describes specific embodiments of the present application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be performed in an order other than that described in the embodiments and still achieve desirable results. In addition, the processes depicted in the figures do not necessarily require the particular order shown or sequential order to achieve the desired results. In certain implementations, multitasking and parallel processing can be advantageous or possible.

[0063] According to another aspect, embodiments of the present application provide an electric energy meter measurement error analysis device. Figure 2 A schematic block diagram of an electric energy meter measurement error analysis device according to an embodiment is shown. It can be understood that the device can be realized by any device, equipment, platform and equipment cluster with computing and processing capability. As shown in the figure, Figure 2As shown, the device comprises: an acquisition unit 200, a first data processing unit 202, a second data processing unit 204, a third data processing unit 206 and a fourth data processing unit 208. The main functions of each component unit are as follows:

[0064] The acquisition unit 200 is configured to acquire feature data of a current electric energy meter; wherein the feature data comprises power grid data, load data, environmental data and operation data;

[0065] The first data processing unit 202 is configured to determine a multi-dimensional feature matrix based on the power grid data, the load data and the environmental data;

[0066] The second data processing unit 204 is configured to input the multi-dimensional feature matrix into a preset neural network model to obtain a first measurement error;

[0067] The third data processing unit 206 is configured to determine a second measurement error based on the operation data;

[0068] The fourth data processing unit 208 is configured to determine a third measurement error based on the first measurement error and the second measurement error.

[0069] In an embodiment of the present application, the power grid data comprises voltage, voltage deviation rate, frequency and sub-harmonic content; the load data comprises current, load rate and power factor; the environmental data comprises environmental temperature and relative humidity; and the operation data comprises actual value of a sampling resistor, actual transformation ratio of a current transformer, actual value of a voltage dividing resistor, quantization error of an ADC chip and actual power grid voltage.

[0070] In an embodiment of the present application, the third data processing unit 206 is configured to perform the following operations:

[0071] determine a first error influence factor based on the actual value of the sampling resistor and the actual transformation ratio of the current transformer;

[0072] determine a second error influence factor based on the actual value of the voltage dividing resistor, the quantization error of the ADC chip and the actual power grid voltage;

[0073] determine the second measurement error based on the first error influence factor and the second error influence factor.

[0074] In an embodiment of the present application, the second measurement error is determined by the following formula:

[0075]

[0076] In the formula, a is the actual value of the sampling resistor, b is the actual transformation ratio of the current transformer, c is the actual value of the voltage dividing resistor, d is the quantization error of the ADC chip, and e is the actual power grid voltage. a first error influencing factor, an actual value of the sampling resistor, a nominal value of the sampling resistor, a temperature coefficient of the sampling resistor, a deviation between the actual temperature and the nominal temperature, an actual transformation ratio of the current transformer, a nominal transformation ratio of the current transformer, a preset current transformer non-linear error, a second error influencing factor, an actual value of the voltage dividing resistor, a nominal value of the voltage dividing resistor, a temperature coefficient of the voltage dividing resistor, a quantization error of the ADC chip, a zero drift error of the ADC chip, an actual measured grid voltage, a gain error of the ADC chip, a second error influencing factor, a second metering error, an inherent error.

[0077] In an embodiment of the present application, the fourth data processing unit 208 is configured to perform the following operations:

[0078] multiplying the first metering error by a first weight and the second metering error by a second weight to obtain a third metering error, wherein the sum of the first weight and the second weight is one, the first weight is used to represent the confidence of the first metering error, and the second weight is used to represent the confidence of the second metering error.

[0079] In an embodiment of the present application, the first weight is determined by the following formula:

[0080]

[0081] wherein, the first weight, a first metering variance, a difference between the model prediction error and the real error in the i-th first metering error sample, an average of all deviations of the first metering error samples, a number of the first metering error samples, a second metering variance, a number of the second metering error samples, a number of the second metering error samples, an average of all deviations of the second metering error sample, an average of all deviations of the second sample.

[0082] In one embodiment of the present application, the device further comprises a fifth data processing unit, which is configured to perform the following operation:

[0083] When the absolute value of the difference between the first metering error and the second metering error is greater than a first preset value, the step of "obtaining the feature data of the current electric energy meter" is re-executed until the number of cycles reaches a second preset value.

[0084] In one embodiment of the present application, the correction weight is determined by the following formula:

[0085]

[0086] In the formula, is the correction weight, is a weight factor of the harmonic scenario, is a harmonic cumulative integral term, is a weight factor of the load scenario, is a load rate at the current time, is a weight factor of the temperature scenario, is a temperature offset cumulative integral term, is a harmonic content at the current time, is a harmonic memory constant, is an ambient temperature, is a historical time, is a current time, is an exponential decay factor, is a temperature decay factor, is a nonlinear coefficient, is a clipping function.

[0087] In one embodiment of the present application, the device further comprises a sixth data processing unit, which is configured to perform the following operation:

[0088] When the number of cycles reaches the second preset value, the third metering error of each time is counted, and the third metering error of each time is averaged to obtain a final result.

[0089] In one embodiment of the present application, the device further comprises a seventh data processing unit, which is configured to perform the following operation:

[0090] When the third metering error is greater than the third preset value, a correction coefficient of the first weight is obtained, the correction coefficient is multiplied by the first weight to obtain a corrected weight, and the first weight in the step of calculating the third metering error by multiplying the first metering error by the first weight and the second metering error by the second weight is replaced by the corrected weight to obtain a fourth metering error.

[0091] According to another aspect, an embodiment also provides a computer readable storage medium having stored thereon a computer program which, when executed in a computer, causes the computer to perform the method described above. Figure 1

[0092] According to another aspect, an embodiment also provides an electronic device comprising a memory and a processor, the memory having stored thereon executable code that, when executed by the processor, implements the method described above. Figure 1

[0093] The various embodiments in the present application are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment mainly explains the difference from other embodiments. Especially, the device embodiments are described simply because they are basically similar to the method embodiments, and the relevant parts can be referred to the part of the method embodiments.

[0094] Those skilled in the art should be aware that, in the above one or more examples, the functions described in the present application can be implemented in hardware, software, firmware or any combination thereof. When implemented in software, the functions can be stored in a computer readable medium or transmitted as one or more instructions or codes on a computer readable medium.

[0095] The above detailed description of the specific implementation of the present application has further explained the purpose, technical solution and beneficial effects of the present application. It should be understood that the above description is only the specific implementation of the present application and is not used to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. made on the basis of the technical solution of the present application shall be included in the protection scope of the present application.​​

Claims

1. A method of analyzing metering errors of an electric energy meter, characterized by, The method comprises: obtaining feature data of a current electric energy meter; wherein the feature data comprises power grid data, load data, environmental data and operation data; determining a multi-dimensional feature matrix based on the power grid data, the load data and the environmental data; inputting the multi-dimensional feature matrix into a preset neural network model to obtain a first metering error; determining a second metering error based on the operation data; determining a third metering error based on the first metering error and the second metering error; the operation data comprises an actual value of a sampling resistor, an actual transformation ratio of a current transformer, an actual value of a voltage dividing resistor, a quantization error of an ADC chip and an actual power grid voltage; the determining of the second metering error based on the operation data comprises: determining a first error influence factor based on the actual value of the sampling resistor and the actual transformation ratio of the current transformer; determining a second error influence factor based on the actual value of the voltage dividing resistor, the quantization error of the ADC chip and the actual power grid voltage; determining the second metering error based on the first error influence factor and the second error influence factor; the second metering error is determined by the following formula: wherein, is the first error influence factor, is the actual value of the sampling resistor, is the nominal value of the sampling resistor, is the temperature coefficient of the sampling resistor, is the deviation of the actual temperature from the nominal temperature, is the actual transformation ratio of the current transformer, is the nominal transformation ratio of the current transformer, is the preset non-linear error of the current transformer, is the second error influence factor, is the actual value of the voltage dividing resistor, is the nominal value of the voltage dividing resistor, is the temperature coefficient of the voltage dividing resistor, is the quantization error of the ADC chip, is the zero drift error of the ADC chip, is the actual measured grid voltage, is the gain error of the ADC chip, is the second error influence factor, is the second metering error, is the inherent error.

2. The method of claim 1, wherein, the power grid data comprises voltage, voltage deviation rate, frequency and sub-harmonic content; the load data comprises current, load rate and power factor; the environmental data comprises environmental temperature and relative humidity; the operation data comprises an actual value of a sampling resistor, an actual transformation ratio of a current transformer, an actual value of a voltage dividing resistor, a quantization error of an ADC chip and an actual power grid voltage.

3. The method of claim 1, wherein, the determining of the third metering error based on the first metering error and the second metering error comprises: multiplying the first metering error by a first weight and the second metering error by a second weight to obtain the third metering error; wherein the sum of the first weight and the second weight is one, the first weight is used to represent the confidence of the first metering error, and the second weight is used to represent the confidence of the second metering error.

4. The method of claim 3, wherein, the first weight is determined by the following formula: wherein, is the first weight, is the first gauge variance, is the first gauge error sample, is the difference between the model prediction error and the true error in the first gauge error sample, is the average of all deviations of the first gauge error sample, is the number of first gauge error samples, is the second gauge variance, is the number of second gauge error samples, is the average of all deviations of the second gauge error sample, is the average of all deviations of the second sample.

5. The method of claim 1, wherein, before the determining of the third metering error based on the first metering error and the second metering error, the method further comprises: when the absolute value of the difference between the first metering error and the second metering error is greater than a first preset value, re-executing the step of obtaining the feature data of the current electric energy meter until the number of cycles reaches a second preset value.

6. An electric energy metering error analysis device, characterized by, The method comprises: an obtaining unit configured to obtain feature data of a current electric energy meter; wherein the feature data comprises power grid data, load data, environmental data and operation data; a first data processing unit configured to determine a multi-dimensional feature matrix based on the power grid data, the load data and the environmental data; a second data processing unit configured to input the multi-dimensional feature matrix into a preset neural network model to obtain a first metering error; a third data processing unit configured to determine a second metering error based on the operation data; a fourth data processing unit configured to determine a third metering error based on the first metering error and the second metering error; The operation data includes an actual value of a sampling resistor, an actual transformation ratio of a current transformer, an actual value of a voltage dividing resistor, a quantization error of an ADC chip, and an actual power grid voltage; The third data processing unit is configured to perform the following operations: determine a first error influence factor based on the actual value of the sampling resistor and the actual transformation ratio of the current transformer; determine a second error influence factor based on the actual value of the voltage dividing resistor, the quantization error of the ADC chip, and the actual power grid voltage; determine the second metering error based on the first error influence factor and the second error influence factor; The second metering error is determined by the following formula: wherein, is the first error influence factor, is the actual value of the sampling resistance, is the nominal value of the sampling resistance, is the temperature coefficient of the sampling resistance, is the deviation of the actual temperature from the nominal temperature, is the actual transformation ratio of the current transformer, is the nominal transformation ratio of the current transformer, is the preset current transformer nonlinearity error, is the second error influence factor, is the actual value of the voltage dividing resistance, is the nominal value of the voltage dividing resistance, is the temperature coefficient of the voltage dividing resistance, is the quantization error of the ADC chip, is the zero drift error of the ADC chip, is the actual measured grid voltage, is the gain error of the ADC chip, is the second error influence factor, is the second metering error, is the inherent error.

7. An electronic device, comprising: A computer program product, comprising a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the method according to any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, A computer program product, comprising a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the method according to any one of claims 1-5.

Images