Electricity meter data calibration and reliability measurement method based on multi-factor dynamic evaluation

Abstract

The present invention discloses an electric energy meter data calibration and reliability measurement method based on multi-factor dynamic evaluation, comprising the following steps: S1, determining the default metering parameter value of the electric energy meter; S2, real-time acquisition of multi-factor data in the electric energy meter and preprocessing, constructing a multi-factor data set; S3, weighted evaluation of the multi-factor data set, generating a storage demand prediction value; S4, dynamically dividing the Flash storage area using a spatiotemporal adaptive topological mapping algorithm to obtain Flash1 and Flash2; S5, writing the corrected metering parameters and CRC check code into the storage unit in the electric energy meter, and synchronously writing them into Flash2; S6, obtaining the correct metering parameters during the power-on startup and normal operation of the electric energy meter; S7, regularly reading back and checking the real-time metering parameters of the metering unit of the electric energy meter. The present invention adopts multi-factor dynamic evaluation and spatiotemporal adaptive topological mapping algorithms to optimize the division of the storage area of ​​the electric energy meter, and has the advantages of efficient storage, accurate data and high reliability.

Description

Technical Field

[0001] The present invention relates to the technical field of electric energy meters, and in particular to an electric energy meter data calibration and reliability measurement method based on multi-factor dynamic evaluation. Background Art

[0002] With the development of smart grid and smart energy metering technology, modern energy meters have been significantly improved in function and performance. Traditional energy meters usually rely on a single measurement and storage solution. The relatively simple metering structure can no longer meet the requirements of modern energy meters for storage and data calibration in complex environments. In order to improve the measurement accuracy and reliability of energy meters, the storage and calibration methods of energy meters are developing in a more intelligent and dynamic direction. Existing technologies mainly rely on fixed storage strategies and simple calibration methods. Although these methods can ensure the basic operation of energy meters, they expose some technical bottlenecks when faced with large-scale, complex and changeable data and storage requirements.

[0003] The traditional storage structure of electric energy meters usually stores all data in a single storage area. This method has certain limitations in storage capacity and storage efficiency. Especially as the functions of electric energy meters increase and the amount of data gradually increases, a single storage area is prone to problems of tight storage resources and insufficient storage space. In addition, traditional electric energy meters often rely on simple static calibration methods, which makes them less adaptable to environmental changes and equipment aging. As the equipment is used for a longer time, the calibration accuracy of the electric energy meter gradually decreases, affecting its long-term reliability.

[0004] In order to deal with these problems, the existing technology attempts to introduce a multi-factor dynamic evaluation method to optimize the division of storage areas and data calibration strategies by real-time collection and analysis of multiple influencing factors. However, these methods often only optimize from one aspect and lack a comprehensive intelligent decision-making mechanism. For example, in the existing technology, the division of storage areas usually relies on simple thresholds or rules, which cannot respond to changes in storage requirements in real time, and it is difficult to achieve accurate optimization allocation. This makes it impossible for the electric energy meter to effectively optimize storage resources in a complex usage environment, and may even cause the storage area to be overcrowded or idle, affecting the long-term stability and data accuracy of the electric energy meter.

[0005] In addition, many existing technologies still rely on traditional calibration methods, which are often based on fixed default parameter values ​​or simple parameter corrections. This approach ignores the dynamic changes of the energy meter at different stages of use and in different environments, making it difficult to achieve personalized and accurate calibration. Especially when the energy meter needs to operate stably for a long time in multiple different environments, traditional calibration methods often cannot adapt to these changes, resulting in fluctuations in measurement accuracy and data reliability.

[0006] In terms of data storage, the Flash storage area used in traditional electric energy meters is usually divided statically. Once the division is completed, the storage area will no longer be dynamically adjusted. This fixed storage area division method fails to fully consider the actual use of storage space and changes in storage demand, which may lead to waste or insufficient storage resources. Especially when the storage demand fluctuates greatly, the storage management system of the electric energy meter may not be able to respond in time, affecting the normal operation of the electric energy meter.

[0007] With the development of smart energy meter technology, the requirements for storage resources and data calibration accuracy are getting higher and higher. The existing static storage division and calibration methods can no longer meet the needs of modern energy meters for efficient and intelligent storage in actual use.

[0008] Therefore, how to provide an electric energy meter data calibration and reliability measurement method based on multi-factor dynamic evaluation is an urgent problem that technicians in this field need to solve. Summary of the invention

[0009] One purpose of the present invention is to propose a method for calibrating and measuring the reliability of electric energy meter data based on multi-factor dynamic evaluation. The present invention intelligently optimizes the division of electric energy meter storage areas through multi-factor dynamic evaluation and spatiotemporal adaptive topological mapping algorithm, and effectively avoids the waste and shortage of storage resources based on accurate storage demand prediction. This technology improves storage efficiency and extends the service life of electric energy meters by comprehensively considering the topological relationship, health status and storage demand of storage blocks. The calibration method combined with the multi-objective genetic algorithm ensures the accuracy and reliability of electric energy meter data, and has the advantages of efficient storage, high data accuracy, strong stability and high reliability.

[0010] According to an embodiment of the present invention, a method for calibrating and measuring reliability of electric energy meter data based on multi-factor dynamic evaluation includes the following steps:

[0011] S1. Determine the default metering parameter value of the electric energy meter and store it in the electric energy meter program;

[0012] S2, real-time collection of multi-factor data in the electric energy meter and pre-processing to construct a multi-factor data set;

[0013] S3, performing a weighted evaluation on the multi-factor data set to generate a storage demand prediction value;

[0014] S4. Based on the predicted value of storage demand, a time-space adaptive topology mapping algorithm is used to dynamically divide the Flash storage area to obtain Flash1 and Flash2, wherein Flash1 is used to store the electric energy meter program, and Flash2 is used to store metering parameters and electric quantity data;

[0015] S5. After the electric energy meter is produced and the error debugging process is performed, the corrected metering parameters and CRC check code are written into the storage unit in the electric energy meter and synchronously written into Flash2;

[0016] S6. During the power-on startup and normal operation of the electric energy meter, the metering parameters are restored from the storage unit of the electric energy meter, and a CRC check is performed. If the check is normal, it means that the correct metering parameters are obtained. If the check is wrong, the metering parameters are restored from Flash2, and a CRC check is performed again. If the check is normal, it means that the correct metering parameters are obtained. If the check fails, the default metering parameter values ​​in the Flash1 electric energy meter program are used;

[0017] S7. Regularly read back and check the real-time metering parameters of the metering unit of the electric energy meter. If they are inconsistent, command the metering unit to reset and write the correct metering parameters into the metering unit.

[0018] Optionally, the default metering parameter value is determined by the hardware design and component specification parameters of the electric energy meter; when the hardware design and component specification parameters remain unchanged, the initial metering error of the electric energy meter remains consistent.

[0019] Optionally, the multi-factor data specifically includes program update frequency, data write frequency, ambient temperature, storage health and MCU load.

[0020] Optionally, the S3 specifically includes:

[0021] S31. For each factor data in the multi-factor data set Normalization is performed using The normalization method obtains the normalization factor value;

[0022] S32. Dynamically adjust the weight of each normalization factor value based on historical data and current workload , the weight satisfy:

[0023] ;

[0024] in, represents the total number of normalization factors, Indicates The weight value of the normalization factor;

[0025] S33, normalize the factor value and weight Combined, calculate the preliminary storage demand forecast:

[0026] ;

[0027] in, represents the preliminary storage demand forecast value, Indicates The normalization factor and The interaction weights between the normalization factors, Indicates The weight of the normalization factor, Indicates Normalization factor value, Indicates Normalization factor value;

[0028] S34, introduce principal component analysis method to predict the initial storage demand Perform dimensionality reduction and extract feature components ;

[0029] S35, using multi-objective genetic algorithm to characterize the components Perform optimization evaluation to generate the final storage demand forecast:

[0030] ;

[0031] in, represents the predicted value of storage demand, Indicates The optimized weights of the principal components, Represents the total number of principal components.

[0032] Optionally, the S4 specifically includes:

[0033] S41. Receive and analyze the generated storage demand forecast value ;

[0034] S42. Constructing the spatiotemporal topology of the Flash storage area ,in Represents a collection of Flash storage blocks. Indicates the connection relationship between Flash storage blocks;

[0035] S43, using the spatiotemporal adaptive topological mapping algorithm, define the spatiotemporal adaptive weight matrix , and calculate the fitness weight:

[0036] ;

[0037] in, Indicates at a point in time Lower storage block With storage block Storage demand forecast The adaptability weight of Represents a storage block With storage block The topological correlation function value between Represents a storage block With storage block The topological correlation function value between Indicates time point The storage demand forecast value, Indicates the total number of nodes;

[0038] S44. Applying spatiotemporal optimization functions Dynamically divide the Flash storage area:

[0039] ;

[0040] in, and Represents the weight coefficient, which is used to balance the satisfaction of storage requirements and the health of storage;

[0041] Represents the storage demand satisfaction function:

[0042] ;

[0043] Represents the storage health function:

[0044] ;

[0045] in, Represents a storage block The number of erase and write times;

[0046] S45. Use Lagrange multiplier method or penalty function method to optimize the space-time function Solve to satisfy the constraints:

[0047] ;

[0048] ;

[0049] in, Represents a storage block storage capacity, Represents a storage block Whether it is selected to store the binary variable of the demand;

[0050] S46, dynamically divide the Flash storage area into Flash1 and Flash2:

[0051] ;

[0052] .

[0053] Optionally, the CRC check is 2-byte data generated by the data involved in the calculation. Before storing the data, the data and the CRC generated by the data are written into a storage unit. When the data is read back from the storage unit for inspection, the check code CRC2 generated by the data is recalculated and compared with CRC and CRC2. If they are equal, the read-back data is considered correct, otherwise the read-back data is considered incorrect.

[0054] Optionally, if the reset in S7 fails to make the metering unit work normally for three consecutive times, the metering unit is considered to be faulty, the current time and power data in the electric energy meter are saved and recorded, and an ERROR fault reminder is displayed.

[0055] Optionally, the electrical structure of the electric energy meter specifically includes an MCU, a metering unit, a storage unit, a communication unit, a display unit, Flash1 and Flash2, and the Flash1 and Flash2 are located in the MCU.

[0056] The beneficial effects of the present invention are:

[0057] First, by performing weighted evaluation on multi-factor data sets, the present invention can more accurately predict the storage requirements of the electric energy meter, thereby optimizing the division of storage areas, avoiding waste or shortage of storage resources, and ensuring the stability and efficiency of the electric energy meter in long-term operation.

[0058] Secondly, the spatiotemporal adaptive topological mapping algorithm adopted by the present invention can dynamically adjust the division of the Flash storage area according to the real-time storage demand prediction value, thereby realizing the intelligent management of the storage area. The algorithm takes into account the topological relationship between the storage blocks and the influence of time changes, and can efficiently optimize the layout of the storage area, balance the storage demand and the health status, thereby improving the storage efficiency and health of the electric energy meter and extending its service life. At the same time, the Flash of the storage program in the MCU is divided into an area Flash2 as the storage space for the metering parameters and core power backup of the electric energy meter, which effectively solves the problem of abnormal metering of the electric energy meter when the storage unit in the electric energy meter fails, and can ensure that the power is not lost.

[0059] In addition, the present invention also introduces a multi-objective genetic algorithm in the data calibration of the electric energy meter. Through the comprehensive evaluation and optimization of multi-factor data, a more accurate storage demand prediction value is generated to ensure that the metering parameters of the electric energy meter are always in the best calibration state. The multi-objective optimization characteristics of the genetic algorithm enable the present invention to find the optimal balance between multiple objectives, effectively improving the measurement accuracy and data accuracy of the electric energy meter. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0061] Figure 1 This is an overall flow chart of the electric energy meter data calibration and reliability measurement method based on multi-factor dynamic evaluation proposed by the present invention;

[0062] Figure 2 The present invention provides a flow chart for obtaining correct calculation parameters for an electric energy meter according to the electric energy meter data calibration and reliability measurement method based on multi-factor dynamic evaluation. DETAILED DESCRIPTION

[0063] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, which only illustrate the basic structure of the present invention in a schematic manner, and therefore only show the components related to the present invention.

[0064] refer to Figure 1 and Figure 2 , an electric energy meter data calibration and reliability measurement method based on multi-factor dynamic evaluation includes the following steps:

[0065] S1. Determine the default metering parameter value of the electric energy meter and store it in the electric energy meter program;

[0066] S2, real-time collection of multi-factor data in the electric energy meter and pre-processing to construct a multi-factor data set;

[0067] S3, performing a weighted evaluation on the multi-factor data set to generate a storage demand prediction value;

[0068] S4. Based on the predicted value of storage demand, a time-space adaptive topology mapping algorithm is used to dynamically divide the Flash storage area to obtain Flash1 and Flash2, wherein Flash1 is used to store the electric energy meter program, and Flash2 is used to store metering parameters and electric quantity data;

[0069] S5. After the electric energy meter is produced and the error debugging process is performed, the corrected metering parameters and CRC check code are written into the storage unit in the electric energy meter and synchronously written into Flash2;

[0070] S6. During the power-on startup and normal operation of the electric energy meter, the metering parameters are restored from the storage unit of the electric energy meter, and a CRC check is performed. If the check is normal, it means that the correct metering parameters are obtained. If the check is wrong, the metering parameters are restored from Flash2, and a CRC check is performed again. If the check is normal, it means that the correct metering parameters are obtained. If the check fails, the default metering parameter values ​​in the Flash1 electric energy meter program are used;

[0071] S7. Regularly read back and check the real-time metering parameters of the metering unit of the electric energy meter. If they are inconsistent, command the metering unit to reset and write the correct metering parameters into the metering unit.

[0072] In this implementation manner, the default metering parameter value is determined by the hardware design and component specification parameters of the electric energy meter; when the hardware design and component specification parameters remain unchanged, the initial metering error of the electric energy meter remains consistent.

[0073] In this implementation, the multi-factor data specifically includes program update frequency, data write frequency, ambient temperature, storage health, and MCU load.

[0074] In this implementation, S3 specifically includes:

[0075] S31. For each factor data in the multi-factor data set Normalization is performed using The normalization method obtains the normalization factor value;

[0076] S32. Dynamically adjust the weight of each normalization factor value based on historical data and current workload , the weight satisfy:

[0077] ;

[0078] in, represents the total number of normalization factors, Indicates The weight value of the normalization factor;

[0079] S33, normalize the factor value and weight Combined, calculate the preliminary storage demand forecast:

[0080] ;

[0081] in, represents the preliminary storage demand forecast value, Indicates The normalization factor and The interaction weights between the normalization factors, Indicates The weight of the normalization factor, Indicates Normalization factor value, Indicates Normalization factor value;

[0082] S34, introduce principal component analysis method to predict the initial storage demand Perform dimensionality reduction and extract feature components ;

[0083] S35, using multi-objective genetic algorithm to characterize the components Perform optimization evaluation to generate the final storage demand forecast:

[0084] ;

[0085] in, represents the predicted value of storage demand, Indicates The optimized weights of the principal components, Represents the total number of principal components.

[0086] In this implementation, S4 specifically includes:

[0087] S41. Receive and analyze the generated storage demand forecast value ;

[0088] S42. Constructing the spatiotemporal topology of the Flash storage area ,in Represents a collection of Flash storage blocks. Indicates the connection relationship between Flash storage blocks;

[0089] S43, using the spatiotemporal adaptive topological mapping algorithm, define the spatiotemporal adaptive weight matrix , and calculate the fitness weight:

[0090] ;

[0091] in, Indicates at a point in time Lower storage block With storage block Storage demand forecast The adaptability weight of Represents a storage block With storage block The topological correlation function value between Represents a storage block With storage block The topological correlation function value between Indicates time point The storage demand forecast value, Indicates the total number of nodes;

[0092] S44. Applying spatiotemporal optimization functions Dynamically divide the Flash storage area:

[0093] ;

[0094] in, and Represents the weight coefficient, which is used to balance the satisfaction of storage requirements and the health of storage;

[0095] Represents the storage demand satisfaction function:

[0096] ;

[0097] Represents the storage health function:

[0098] ;

[0099] in, Represents a storage block The number of erase and write times;

[0100] S45. Use Lagrange multiplier method or penalty function method to optimize the space-time function Solve to satisfy the constraints:

[0101] ;

[0102] ;

[0103] in, Represents a storage block storage capacity, Represents a storage block Whether it is selected to store the binary variable of the demand;

[0104] S46, dynamically divide the Flash storage area into Flash1 and Flash2:

[0105] ;

[0106] .

[0107] In this embodiment, the CRC check is a 2-byte data generated by the data involved in the calculation. Before storing the data, the data and the CRC generated by the data are written into the storage unit. When the data is read back from the storage unit for inspection, the check code CRC2 generated by the data is recalculated and compared with CRC and CRC2. If they are equal, the read-back data is considered correct, otherwise the read-back data is considered incorrect.

[0108] In this embodiment, if the reset in S7 fails to make the metering unit work normally for three consecutive times, it is considered that the metering unit is faulty, the current time and power data in the electric energy meter are saved and recorded, and an ERROR fault reminder is displayed.

[0109] In this implementation, the electrical structure of the electric energy meter specifically includes an MCU, a metering unit, a storage unit, a communication unit, a display unit, Flash1 and Flash2, and the Flash1 and Flash2 are located in the MCU.

[0110] Embodiment 1:

[0111] In order to verify the feasibility of the present invention in implementation, the present invention is applied to the intelligent power metering system of a certain city. The system covers multiple high-density power consumption areas, with a large number of power meters, and with the fluctuation of power demand, the data storage requirements of the power meters will also fluctuate greatly. In this environment, the traditional way of dividing the storage area of ​​the power meter can no longer meet the demand, especially when the power meter needs to store more metering data and parameters, there is often a lack of storage space.

[0112] In this embodiment, the method of the present invention first collects multi-factor data (such as program update frequency, data write frequency, ambient temperature, etc.) in the electric energy meter and pre-processes these data to construct a multi-factor data set. Then, based on the data set, the storage demand prediction value of the electric energy meter is calculated through weighted evaluation. Next, the Flash storage area is dynamically divided using the spatiotemporal adaptive topology mapping algorithm, and Flash1 is used to store the electric energy meter program, while Flash2 is used to store metering parameters and core power data. This division not only improves the storage efficiency, but also ensures the high reliability of the metering data.

[0113] In actual application, we used 50 smart energy meters in high-density electricity consumption areas in a city for testing. These energy meters operate in different environments, including different temperatures, humidity and different electricity consumption modes. During data collection, we monitored multiple factors of these energy meters and dynamically divided the storage areas using the method of the present invention.

[0114] During the implementation process, we first collected and preprocessed the data. After normalizing the multi-factor data of the energy meter, we calculated the storage demand forecast based on the weight evaluation algorithm, and dynamically divided the Flash storage area of ​​the energy meter using the spatiotemporal adaptive topology mapping algorithm based on the forecast value. Then, we optimized the division of the storage area through the graph neural network to ensure the efficient use of the storage space.

[0115] Through the optimized storage structure of the energy meter, we can more accurately control the storage requirements of the energy meter and ensure the storage reliability under different workloads. At the same time, the energy meter can realize real-time measurement parameter calibration and storage data readback check during operation, which improves the accuracy and reliability of the data.

[0116] Table 1 Performance comparison of energy meter storage solutions based on multi-factor dynamic evaluation optimization

[0117]

[0118] Based on the data provided in Table 1 above, the performance of the electric energy meters in each monitoring area can be analyzed in detail.

[0119] First, the metering error rate is a key indicator for measuring the accuracy of the electric energy meter. The error rates of all monitoring areas are kept within 1.15%. Among them, the metering error rate of area A is the lowest, only 0.95%, indicating that the electric energy meters in this area have high metering accuracy. The metering error rate of area B is 1.10%, slightly higher than that of area A, but still within a reasonable range. The metering error rates of areas C, D, and E are 1.05%, 1.00%, and 1.15%, respectively, with small differences, showing that the stability of the electric energy meters in this area is at the same level.

[0120] Secondly, the system failure rate reflects the stability and reliability of the electric energy meter. As can be seen from Table 1, the system failure rate in area C is the lowest, only 0.01%, which means that the failure rate of the electric energy meters in this area is extremely low and the reliability is high. The system failure rates in areas D and E are also low, at 0.01% and 0.02%, respectively. In contrast, the system failure rates in areas A and B are slightly higher, at 0.02% and 0.03%, respectively, but the overall failure rate is still within an acceptable range, indicating that the electric energy meters in each area perform well in terms of stability.

[0121] The average response time is an important indicator to measure the response speed of the system. According to the data, the energy meter in area B has the fastest response time, with an average response time of 18 milliseconds, indicating that the energy meter in this area responds more quickly to the input signal. It is followed by area E, with a response time of 19 milliseconds. The response times of areas A and D are 20 milliseconds and 21 milliseconds, which are slightly worse than those of areas B and E, but the difference is not large. The response time of area C is 22 milliseconds, which is the longest response time among all areas, which may be related to the network delay or system configuration in this area.

[0122] Finally, the power data loss rate is a key indicator for evaluating the data integrity and system stability of the energy meter. The power data loss rate in all areas remained at an extremely low level, with the minimum being 0.0005% in area C and the maximum being 0.002% in area E. This shows that no matter in which area, the energy meter can effectively avoid data loss and ensure data reliability during normal operation. In particular, the power data loss rate in area C is the lowest, indicating that the energy meter in this area has excellent performance in data transmission and storage.

[0123] In summary, the data in Table 1 show that the electric energy meters in all monitoring areas perform well in terms of metering accuracy, system failure rate, response time, and data loss rate. Although there are some minor differences, these differences are generally within an acceptable range and indicate that the performance of the electric energy meters in different areas can meet the needs of daily use. After applying the multi-factor dynamic evaluation method of the present invention, the performance of the electric energy meters in each area has been significantly improved, which can better meet the requirements of power metering and data integrity, thereby ensuring the stability and accuracy of the electric energy meters in actual operation.

[0124] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. The electric energy meter data calibration and reliability measurement method based on multi-factor dynamic evaluation is characterized by: The steps include: S1. Determine the default metering parameter value of the electric energy meter and store it in the electric energy meter program; S2, real-time collection of multi-factor data in the electric energy meter and pre-processing to construct a multi-factor data set; S3, performing a weighted evaluation on the multi-factor data set to generate a storage demand prediction value; S4. Based on the predicted value of storage demand, a time-space adaptive topology mapping algorithm is used to dynamically divide the Flash storage area to obtain Flash1 and Flash2, wherein Flash1 is used to store the electric energy meter program, and Flash2 is used to store metering parameters and electric quantity data; S5. After the electric energy meter is produced and the error debugging process is performed, the corrected metering parameters and CRC check code are written into the storage unit in the electric energy meter and synchronously written into Flash2; S6. During the power-on startup and normal operation of the electric energy meter, the metering parameters are restored from the storage unit of the electric energy meter, and a CRC check is performed. If the check is normal, it means that the correct metering parameters are obtained. If the check is wrong, the metering parameters are restored from Flash2, and a CRC check is performed again. If the check is normal, it means that the correct metering parameters are obtained. If the check fails, the default metering parameter values ​​in the Flash1 electric energy meter program are used; S7, regularly read back and check the real-time metering parameters of the metering unit of the electric energy meter. If they are inconsistent, command the metering unit to reset and write the correct metering parameters into the metering unit; The multi-factor data specifically includes program update frequency, data writing frequency, ambient temperature, storage health and MCU load; The S3 specifically includes: S31. For each factor data in the multi-factor data set Normalization is performed using The normalization method obtains the normalization factor value; S32. Dynamically adjust the weight of each normalization factor value based on historical data and current workload , the weight satisfy: ； in, represents the total number of normalization factors, Indicates The weight value of the normalization factor; S33, normalize the factor value and weight Combined, calculate the preliminary storage demand forecast: ； in, represents the preliminary storage demand forecast value, Indicates The normalization factor and The interaction weights between the normalization factors, Indicates The weight of the normalization factor, Indicates Normalization factor value, Indicates Normalization factor value; S34, introduce principal component analysis method to predict the initial storage demand Perform dimensionality reduction and extract feature components ; S35, using multi-objective genetic algorithm to characterize the components Perform optimization evaluation to generate the final storage demand forecast: ； in, represents the predicted value of storage demand, Indicates The optimized weights of the principal components, Represents the total number of principal components.

2. The method for calibrating and measuring reliability of electric energy meter data based on multi-factor dynamic evaluation according to claim 1 is characterized in that: The default metering parameter value is determined by the hardware design and component specification parameters of the electric energy meter; when the hardware design and component specification parameters remain unchanged, the initial metering error of the electric energy meter remains consistent.

3. The method for calibrating and measuring reliability of electric energy meter data based on multi-factor dynamic evaluation according to claim 1 is characterized in that: The S4 specifically includes: S41. Receive and analyze the generated storage demand forecast value ; S42. Constructing the spatiotemporal topology of the Flash storage area ,in Represents a collection of Flash storage blocks. Indicates the connection relationship between Flash storage blocks; S43, using the spatiotemporal adaptive topological mapping algorithm, define the spatiotemporal adaptive weight matrix , and calculate the fitness weight: ； in, Indicates at a point in time Lower storage block With storage block Storage demand forecast The adaptability weight of Represents a storage block With storage block The topological correlation function value between Represents a storage block With storage block The topological correlation function value between Indicates time point The storage demand forecast value, Indicates the total number of nodes; S44. Applying spatiotemporal optimization functions Dynamically divide the Flash storage area: ； in, and Represents the weight coefficient, which is used to balance storage demand satisfaction and storage health; Represents the storage demand satisfaction function: ； Represents the storage health function: ； Among them, the number of erase and write times Represents a storage block The number of erase and write times; S45. Use Lagrange multiplier method or penalty function method to optimize the space-time function Solve to satisfy the constraints: ； ； Among them, capacity Represents a storage block storage capacity, Represents a storage block Whether it is selected to store the binary variable of the demand; S46, dynamically divide the Flash storage area into Flash1 and Flash2: ； 。 4. The method for calibrating and measuring reliability of electric energy meter data based on multi-factor dynamic evaluation according to claim 1 is characterized in that: The CRC check is a 2-byte data generated by the data involved in the calculation. Before storing the data, the data and the CRC generated by the data are written into the storage unit. When the data is read back from the storage unit for inspection, the check code CRC2 generated by the data is recalculated and compared with CRC and CRC2. If they are equal, the read-back data is considered correct, otherwise the read-back data is considered incorrect.

5. The method for calibrating and measuring reliability of electric energy meter data based on multi-factor dynamic evaluation according to claim 1 is characterized in that: If the reset in S7 fails to make the metering unit work normally for three consecutive times, it is considered that the metering unit is faulty, the current time and power data in the electric energy meter are saved and recorded, and an ERROR fault reminder is displayed.

6. The method for calibrating and measuring reliability of electric energy meter data based on multi-factor dynamic evaluation according to claim 1 is characterized in that: The electrical structure of the electric energy meter specifically includes an MCU, a metering unit, a storage unit, a communication unit, a display unit, Flash1 and Flash2, and the Flash1 and Flash2 are located in the MCU.

Images