Extra-high voltage direct current transmission electric energy metering method, device and equipment

Abstract

The invention discloses an extra-high-voltage direct-current transmission electric energy metering method, device and equipment, relates to the technical field of extra-high-voltage direct-current transmission, and can solve the technical problem of low electric energy metering accuracy in the current extra-high-voltage direct-current transmission scene. Comprising the following steps: determining a plurality of parameter items for electric energy metering, and obtaining historical parameter values corresponding to the parameter items; obtaining an electric energy metering historical actual value, training an initial multiple regression error calculation model according to the historical parameter value and the electric energy metering historical actual value, and obtaining a trained multiple regression error calculation model; and obtaining a real-time parameter value corresponding to the parameter item, performing error prediction on the real-time parameter value by using the multiple regression error calculation model to obtain an error value, obtaining an electric energy metering predicted value, and obtaining an electric energy metering corrected value according to the error value and the electric energy metering predicted value.

Description

Technical Field

[0001] The present invention relates to the technical field of ultra-high voltage direct current transmission, and in particular to an ultra-high voltage direct current transmission electric energy metering method, device and equipment. Background Art

[0002] Ultra-high voltage direct current (UHVDC) technology is a key technology for achieving long-distance, large-capacity power transmission, and can effectively solve the problems of uneven distribution of power resources and grid stability. The efficient operation of UHVDC systems places higher demands on the accuracy and reliability of power metering. With the increase in power demand and the increase in grid complexity, traditional power metering methods face many challenges in the UHVDC environment.

[0003] Traditional metering methods mainly rely on current and voltage sensors to calculate electric energy by measuring the instantaneous values ​​of current and voltage. However, this method is easily interfered by external factors in complex environments, resulting in increased metering errors. For example, changes in ambient temperature and humidity can affect the measurement accuracy of sensors, and the length and transmission distance of transmission lines can also cause fluctuations in voltage and current, thereby affecting the metering results. Therefore, there is an urgent need to improve the accuracy of UHV DC transmission energy metering. Summary of the invention

[0004] In view of this, the present invention provides a method, device and equipment for ultra-high voltage direct current transmission electric energy metering, which can solve the technical problem of low accuracy of electric energy metering in the current ultra-high voltage direct current transmission scenario.

[0005] According to a first aspect of the present invention, there is provided a method for measuring electric energy of ultra-high voltage direct current transmission, which is applied to an edge computing unit, and the method comprises:

[0006] Determine a plurality of parameter items for electric energy metering, and obtain historical parameter values ​​corresponding to the parameter items;

[0007] Acquire historical actual values ​​of electric energy metering, train an initial multiple regression error calculation model according to the historical parameter values ​​and the historical actual values ​​of electric energy metering, and obtain a trained multiple regression error calculation model;

[0008] Obtain the real-time parameter value corresponding to the parameter item, use the multivariate regression error calculation model to perform error prediction on the real-time parameter value to obtain the error value, obtain the electric energy metering prediction value, and obtain the electric energy metering correction value according to the error value and the electric energy metering prediction value.

[0009] According to a second aspect of the present invention, there is provided an ultra-high voltage direct current transmission power metering device, which is applied to an edge computing unit, and the device comprises:

[0010] A determination module, used to determine a plurality of parameter items for electric energy metering, and obtain historical parameter values ​​corresponding to the parameter items;

[0011] A training module is used to obtain the historical actual values ​​of electric energy metering, train an initial multiple regression error calculation model according to the historical parameter values ​​and the historical actual values ​​of electric energy metering, and obtain a trained multiple regression error calculation model;

[0012] The correction module is used to obtain the real-time parameter value corresponding to the parameter item, use the multivariate regression error calculation model to perform error prediction on the real-time parameter value to obtain the error value, obtain the electric energy metering prediction value, and obtain the electric energy metering correction value according to the error value and the electric energy metering prediction value.

[0013] According to a third aspect of the present application, a storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned ultra-high voltage direct current transmission electric energy metering method is implemented.

[0014] According to the fourth aspect of the present application, a computer device is provided, including a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor, wherein the processor implements the above-mentioned UHV DC transmission electric energy metering method when executing the program.

[0015] Through the above technical scheme, the present invention provides a UHV DC transmission electric energy metering method, device and equipment, which are applied to an edge computing unit. First, multiple parameter items for electric energy metering are determined, and historical parameter values ​​corresponding to the parameter items are obtained; then, the actual historical values ​​of electric energy metering are obtained, and an initial multivariate regression error calculation model is trained according to the historical parameter values ​​and the actual historical values ​​of electric energy metering to obtain a trained multivariate regression error calculation model; finally, the real-time parameter values ​​corresponding to the parameter items are obtained, and the real-time parameter values ​​are predicted by using the multivariate regression error calculation model to obtain error values, obtain electric energy metering prediction values, and obtain electric energy metering correction values ​​according to the error values ​​and the electric energy metering prediction values. Through the technical solution of the present invention, edge computing units are deployed at key nodes, and the computing power of the edge computing units is used to perform error compensation and data processing tasks in real time, reduce data transmission delays, and improve system response speed. Multiple parameter items that affect electric energy metering are obtained, covering all key parameters of the ultra-high voltage transmission system, ensuring the comprehensiveness and accuracy of the data. By training the initial multivariate regression error calculation model, an accurate relationship model between each parameter item and the error of electric energy metering, namely, the multivariate regression error calculation model, is established to ensure the accuracy and stability of the model. The process of using the multivariate regression error calculation model to predict the error of the real-time parameter values ​​corresponding to multiple parameter items is to use the regression coefficient as the weighting coefficient, and use the regression coefficient to fuse the multiple parameter items. The fused data is the error value. This error value is used for correction to ensure the high accuracy and real-time performance of the electric energy metering correction value.

[0016] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented according to the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation on the local application. In the drawings:

[0018] Figure 1 A schematic diagram of a flow chart of a method for measuring electric energy in ultra-high voltage direct current transmission provided by an embodiment of the present invention is shown;

[0019] Figure 2 A schematic flow chart of another method for measuring electric energy of ultra-high voltage direct current transmission provided by an embodiment of the present invention is shown;

[0020] Figure 3 A schematic structural diagram of a UHV DC transmission electric energy metering device provided by an embodiment of the present invention is shown;

[0021] Figure 4 A schematic structural diagram of another ultra-high voltage direct current transmission electric energy metering device provided in an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0022] The present invention will be described in detail below with reference to the accompanying drawings and in combination with embodiments. It should be noted that the embodiments and features in the embodiments of the present invention can be combined with each other without conflict.

[0023] This embodiment provides a method for measuring electric energy in ultra-high voltage direct current transmission, which is applied to an edge computing unit, such as Figure 1 As shown, the method includes:

[0024] 101. Determine a plurality of parameter items for electric energy metering, and obtain historical parameter values ​​corresponding to the parameter items.

[0025] For this embodiment, the multiple parameter items are at least one parameter item, for example, the parameter items include: current, voltage, temperature, humidity and transmission distance, etc. Specifically, an intelligent data acquisition module is deployed at each sensor node, integrating multi-sensor data fusion algorithm, real-time data processing and edge computing capabilities. The intelligent data acquisition module can process and analyze sensor data in real time, perform error correction and data cleaning during preliminary measurement, and connect to a central processing unit or a cloud data center via a high-speed industrial Ethernet or a low-power wide area network (LPWAN).

[0026] Among them, the current sensor: uses an ultra-high precision Hall effect sensor to monitor instantaneous current in real time to ensure stable performance in extreme environments. The current sensor converts analog signals into high-precision digital signals through an intelligent data acquisition module and transmits them in real time.

[0027] Among them, the voltage sensor: uses a high-precision voltage divider and a precision amplifier to obtain voltage in real time, integrates temperature compensation function to ensure measurement accuracy under high voltage conditions, and the output signal of the voltage sensor is processed by a high sampling rate analog-to-digital converter (ADC) and digitally processed by an intelligent data acquisition module to ensure data accuracy and real-time performance.

[0028] Among them, the multi-point temperature sensor array: deploy multiple high-sensitivity temperature sensors at the public node (PPC), the converter valve side of the converter station, the grid side of the converter transformer, and each feeder outlet, etc., and quickly respond to temperature changes through the wireless transmission module. The multi-point temperature sensor array transmits temperature data to the intelligent data acquisition module for real-time analysis and processing to ensure the accuracy of temperature data.

[0029] Among them, the intelligent humidity sensor: uses a micro capacitive humidity sensor with a built-in intelligent calibration algorithm, which can adapt to environmental changes and ensure the accuracy of humidity measurement. Humidity data is transmitted to the intelligent data acquisition module in real time through low-power wireless transmission technology for data fusion and error analysis.

[0030] Among them, the laser ranging transmission distance measuring device: uses laser ranging technology to accurately measure the actual distance of power transmission. The laser ranging transmission distance measuring device uses an intelligent data acquisition module to process data and correct errors during measurement, and combines multi-point ranging data to eliminate errors during measurement.

[0031] In summary, through multiple sensors and intelligent data acquisition modules, multi-dimensional and high-precision data acquisition is achieved. The acquired data is the historical parameter value corresponding to the parameter item, specifically, the historical parameter value corresponding to the parameter item at each historical moment. For example, if there are 5 parameter items and 10 historical moments, the historical parameter values ​​corresponding to the parameter items are the 5 historical parameter values ​​corresponding to the 5 parameter items at historical moment 1 (each parameter item corresponds to a historical parameter value at a historical moment), the 5 historical parameter values ​​corresponding to the 5 parameter items at historical moment 2, ... the 5 historical parameter values ​​corresponding to the 5 parameter items at historical moment 10.

[0032] Existing electricity metering methods often only consider a single or a few parameters, such as current or voltage, and ignore the combined influence of multiple factors such as temperature, humidity, and transmission distance. They are unable to fully reflect the actual situation, resulting in large metering errors and affecting the accuracy of metering.

[0033] 102. Obtain historical actual values ​​of electric energy metering, and train an initial multiple regression error calculation model according to the historical parameter values ​​and the historical actual values ​​of electric energy metering to obtain a trained multiple regression error calculation model.

[0034] For this embodiment, each historical moment corresponds to an actual historical value of electric energy metering. For each historical moment, the actual historical value of electric energy metering and the historical parameter values ​​corresponding to multiple parameter items are taken as a group of data. Multiple historical moments correspond to multiple groups of data. The initial multiple regression error calculation model is trained using the multiple groups of data to obtain a trained multiple regression error calculation model.

[0035] Among them, the multivariate regression error calculation model is constructed based on the multivariate regression analysis method, which is a relationship model between all parameter items and the error of electric energy metering.

[0036] 103. Obtain the real-time parameter value corresponding to the parameter item, use the multivariate regression error calculation model to perform error prediction on the real-time parameter value to obtain the error value, obtain the electric energy metering prediction value, and obtain the electric energy metering correction value according to the error value and the electric energy metering prediction value.

[0037] After the multivariate regression error calculation model is obtained in step 102 of the embodiment, for this embodiment, the multivariate regression error calculation model can be applied, and the real-time parameter values ​​corresponding to each parameter item can be input to obtain the error value. The electric energy metering prediction value obtained by the prior art has a large error, and this error value is used to correct the electric energy metering prediction value. Therefore, the electric energy metering correction value can be obtained according to the error value and the electric energy metering prediction value.

[0038] The present invention provides an ultra-high voltage direct current transmission electric energy metering method, device and equipment, which are applied to an edge computing unit. First, multiple parameter items for electric energy metering are determined, and historical parameter values ​​corresponding to the parameter items are obtained; then, the actual historical values ​​of electric energy metering are obtained, and an initial multivariate regression error calculation model is trained according to the historical parameter values ​​and the actual historical values ​​of electric energy metering to obtain a trained multivariate regression error calculation model; finally, the real-time parameter values ​​corresponding to the parameter items are obtained, and the real-time parameter values ​​are predicted by using the multivariate regression error calculation model to obtain error values, and electric energy metering prediction values ​​are obtained, and electric energy metering correction values ​​are obtained according to the error values ​​and the electric energy metering prediction values. Through the technical solution of the present invention, edge computing units are deployed at key nodes, and the computing power of the edge computing units is used to perform error compensation and data processing tasks in real time, reduce data transmission delays, and improve system response speed. Multiple parameter items that affect electric energy metering are obtained, covering all key parameters of the ultra-high voltage transmission system, ensuring the comprehensiveness and accuracy of the data. By training the initial multivariate regression error calculation model, an accurate relationship model between each parameter item and the error of electric energy metering, namely, the multivariate regression error calculation model, is established to ensure the accuracy and stability of the model. The process of using the multivariate regression error calculation model to predict the error of the real-time parameter values ​​corresponding to multiple parameter items is to use the regression coefficient as the weighting coefficient, and use the regression coefficient to fuse the multiple parameter items. The fused data is the error value. This error value is used for correction to ensure the high accuracy and real-time performance of the electric energy metering correction value.

[0039] Further, as a refinement and expansion of the specific implementation of the above embodiment, in order to fully illustrate the specific implementation process of this embodiment, another ultra-high voltage direct current transmission power metering method is provided, which is applied to the edge computing unit, such as Figure 2 As shown, the method includes:

[0040] It should be noted that the background is that a new UHV transmission project in a certain place needs to build an energy metering system simultaneously to ensure the accuracy and stability of energy metering. During operation, if there is a metering error caused by environmental factors or equipment failure, it needs to be adjusted and compensated in time.

[0041] In the traditional method, each project needs to build a separate metering point after construction, and use traditional data collection and monitoring methods to continuously pay attention to operating status indicators such as voltage, current and power. According to the specific situation of the project, the electric energy metering model is manually built. The traditional state monitoring model of the AC transmission system is usually adopted, and necessary modifications and adjustments are made to meet the requirements of the ultra-high voltage transmission system. Use traditional current transformers, voltage transformers and other sensors to collect data such as voltage, current and power in real time. These data are transmitted to the central processing unit through the data acquisition system. Continue to pay attention to the table code data such as voltage, current, and power to ensure that the transmission system operates within the normal range. Call previous experience models to analyze the collected data. Commonly used algorithms include typical algorithm models such as three-phase imbalance, overvoltage, and overcurrent. These algorithms are based on historical data and empirical rules, set certain thresholds, and judge whether there is an abnormality in the system. When the monitoring data exceeds the set threshold, the system will issue an alarm signal. The alarm signal is usually sent to relevant staff through SMS, email or other communication methods, and then the staff conducts relevant investigations. The traditional method has the following defects:

[0042] 1. The modeling steps are complicated: the model needs to be manually adjusted according to the specific project, which is time-consuming, labor-intensive and prone to errors.

[0043] 2. Incomplete monitoring parameters: Traditional methods only focus on basic parameters such as voltage, current and power, and do not consider the impact of nonlinear factors in the AC / DC conversion process such as rectification, inversion, and environmental factors on the metering system, resulting in incomplete monitoring.

[0044] 3. Inaccurate model algorithm: Traditional methods rely on empirical data or manually set threshold data. These models cannot actually reflect the measurement performance or related risks and are prone to misjudgment or omission.

[0045] 4. Lack of precise positioning function: Traditional methods cannot accurately locate abnormalities or fault points. Manual inspection is still required after receiving the alarm information, which is time-consuming and affects the operating efficiency and reliability of the system.

[0046] Traditional error analysis models are mostly static models and lack dynamic adjustment capabilities. When facing changes in environmental conditions and load fluctuations, the error model cannot be adjusted in real time, which makes it difficult for the metering system to maintain high precision in actual operation, resulting in error accumulation and inability to accurately reflect the actual power consumption. However, the present application, through the embodiment steps 201-208, compared with the traditional method, firstly, realizes one-click system construction and real-time monitoring by automatically calling the model and intelligent data processing, reducing manual intervention. Second, the multi-parameter comprehensive error analysis (i.e., the multivariate regression error calculation model) takes into account a variety of influencing factors, covers all key parameters of the ultra-high voltage transmission system, and ensures the comprehensiveness and accuracy of the data. Third, through adaptive adjustment and incremental learning, accurate positioning and real-time compensation of the metering error are achieved. Fourth, based on the prediction ability and risk rating of the model, advance warning is achieved, fault handling time is reduced, and the stability and reliability of the system are improved. Therefore, the present application significantly improves the accuracy and reliability of the electric energy metering of the ultra-high voltage transmission system, ensures the stable operation of the electric energy metering device in various complex environments, can effectively solve the error problem in the electric energy metering of the ultra-high voltage direct current transmission system, and provides a solid guarantee for fair trading in the power market and the management of the power grid.

[0047] In addition, the embodiment steps 201-208 use edge computing and intelligent gateways to improve data processing efficiency and system response speed, reduce data transmission delay, and ensure efficient operation of the system in an ultra-high voltage direct current transmission environment by processing and transmitting data in real time. Its significance lies in reducing the load of cloud data centers and improving the real-time performance and reliability of the system.

[0048] Edge computing moves data processing and analysis tasks closer to the data source, reducing data transmission delays, improving real-time performance and response speed, and ensuring stable operation of the system in complex environments. The first is to deploy edge computing units at key nodes, such as UHV DC transmission converter stations and important transmission line nodes. Equipped with high-performance embedded processors, storage units, and communication modules, it supports complex data processing and high-speed data transmission. Secondly, the functions of steps 201-208 are integrated in the edge computing unit, and the processed data is stored locally, and a cache is set up to ensure that data is not lost when communication is interrupted or delayed.

[0049] The smart gateway aggregates and transmits data to ensure real-time synchronization and secure transmission of data, reduce cloud load, and enhance the overall efficiency of the system. The smart gateway needs to aggregate data from multiple edge computing units, compress the aggregated data, encrypt it, and transmit it through a high-speed network (5G). Subsequently, the data is synchronized to the cloud data center to ensure data consistency, and then distributed to other edge computing units or local servers to reduce cloud load.

[0050] 201. Determine multiple parameter items for electric energy metering, and obtain historical parameter values ​​corresponding to the parameter items.

[0051] For this embodiment, the determination of multiple parameter items for electric energy metering includes: obtaining a first preset number of original parameter items for electric energy metering and historical original parameter values ​​corresponding to the original parameter items; preprocessing the historical original parameter values ​​to obtain preprocessed historical original parameter values; obtaining electric energy metering historical actual values ​​and corresponding electric energy metering historical predicted values, and calculating the electric energy metering error based on the electric energy metering historical actual values ​​and the electric energy metering historical predicted values; calculating the correlation between the preprocessed historical original parameter values ​​and the electric energy metering error, sorting the correlation, and determining the first second preset number of original parameter items as parameter items for electric energy metering.

[0052] Not all parameters that affect the error in electric energy metering are used as parameter items. Instead, parameters that have a high correlation with the error in electric energy metering are selected from a first preset number of original parameter items that affect the error in electric energy metering as parameter items.

[0053] The historical original parameter value corresponding to the original parameter item refers to the historical original parameter value corresponding to the original parameter item at each historical moment. For example, there are 10 original parameter items and 10 historical moments. The historical original parameter values ​​corresponding to the original parameter items are the 10 historical original parameter values ​​corresponding to the 10 original parameter items at historical moment 1 (each original parameter item corresponds to a historical original parameter value at a historical moment), the 10 historical original parameter values ​​corresponding to the 10 original parameter items at historical moment 2, ... the 10 historical original parameter values ​​corresponding to the 10 original parameter items at historical moment 10.

[0054] Among them, the historical original parameter values ​​are preprocessed, and the preprocessing includes using the average moving method to remove the background noise in the historical original parameter values ​​to obtain smoothed historical original parameter values, and then using the maximum-minimum normalization method to scale the smoothed historical original parameter values ​​to a specific range (usually greater than 0 and less than 1) to obtain target data, and finally using the linear interpolation method to complete the data to obtain the preprocessed historical original parameter values ​​to ensure the quality and consistency of the data.

[0055] Moving average method: Smooth the data and reduce noise by calculating the average value of the data within a certain window.

[0056]

[0057] b(t) is the smoothed historical original parameter value at historical time t, a(ti) is the historical original parameter value at historical time ti, and N is the window size.

[0058] Min-max normalization: This method scales the data to a specific range to facilitate quantitative analysis of the data.

[0059]

[0060] c is the target data, b is the smoothed historical original parameter value, b min and b max are the minimum and maximum values ​​of the data set of the smoothed historical original parameter values, respectively.

[0061] Linear interpolation: Use the linear relationship between adjacent data points to interpolate and fill in missing values.

[0062]

[0063] c(t) is the interpolated data at historical time t, t1 and t2 are the historical times corresponding to adjacent non-missing data, and c(t1) and c(t2) are the corresponding target data.

[0064] After adding c(t) to the target data, the historical original parameter value after preprocessing is obtained.

[0065] Among them, each historical moment corresponds to an actual historical value of electric energy metering and a predicted historical value of electric energy metering. The difference between the two is that before the historical moment, the historical predicted value of electric energy metering is calculated by the existing technology for the historical moment, and after the historical moment, the actual historical value of electric energy metering can be directly obtained for the historical moment.

[0066] Calculating the electric energy metering error according to the electric energy metering historical actual value and the electric energy metering historical predicted value includes: electric energy metering error=|electric energy metering historical actual value-electric energy metering historical predicted value|.

[0067] As an implementation method for calculating the correlation between the historical original parameter value after the preprocessing and the electric energy metering error, the Pearson correlation coefficient is used. The Pearson correlation coefficient r is used to measure the linear relationship between two variables, and its value range is [-1, 1]: r = 1 indicates a completely positive linear correlation; r = -1 indicates a completely negative linear correlation; r = 0 indicates no linear correlation.

[0068] Specifically, for each original parameter item, at each historical moment i there is a corresponding preprocessed historical original parameter value z i and an energy measurement error s i , calculate the historical original parameter value z after all preprocessing i The average Calculate all energy metering errors s i The average Calculate the correlation using the Pearson correlation coefficient:

[0069]

[0070] 202. Obtain historical actual values ​​of electric energy metering, construct regression coefficient terms, construct 1 and the product of the historical parameter values ​​and the corresponding regression coefficient terms, add the products to obtain a model prediction value, construct a difference between the model prediction value and the historical actual values ​​of electric energy metering, and use the square of the difference as an initial multiple regression error calculation model.

[0071] For this embodiment, the regression coefficients are β0, β1, ..., β p , the regression coefficient term corresponding to 1 is β0, and the historical parameter value x of the parameter term x1 is i1 The corresponding regression coefficient term is β1, and the historical parameter value x of the parameter term x2 is i2 The corresponding regression coefficient term is β2, ... parameter term x p The historical parameter value x ip The corresponding regression coefficient term is β p .

[0072] Construct 1 and the product of the historical parameter value and the corresponding regression coefficient term, add the products, and obtain the model prediction value, that is: The actual historical value of electric energy measurement is y i , construct the difference between the model prediction value and the actual value of the electric energy metering history, and use the square of the difference as the initial multiple regression error calculation model, that is,

[0073]

[0074] 203. Calculate, according to the historical parameter values ​​and the actual historical values ​​of the electric energy metering, the initial regression coefficient of each regression coefficient item when the initial multiple regression error calculation model is minimized.

[0075] For this embodiment,

[0076]

[0077] Here, i is a historical moment.

[0078] The initial regression coefficients can be calculated by matrix operations. Let η be the design matrix of n×(p+1), where each row is a vector [1, x i1 , x i2 , ..., x ip ]; y is an n×1 vector, using [y1, y2, ..., y n] is the transposed representation; β is a (p+1)×1 vector, using [β0, β1, ..., β p ] is the transposed representation. The least squares estimate of the initial regression coefficient is as follows:

[0079]

[0080] for

[0081] 204. Test the initial regression coefficient. If the test passes, determine the initial regression coefficient as the target regression coefficient. If the test fails, continue training until the test passes.

[0082] For this embodiment, the testing of the initial regression coefficient includes: testing the initial regression coefficient by using K-fold cross validation and residual analysis.

[0083] K-fold cross validation: Divide the training set into K subsets, use K-1 subsets to train the model each time, and use the remaining subset to validate the model. Repeat K times, calculate the error of each validation, and take the average as the validation error of the model.

[0084]

[0085] Among them, error j is the error of the j-th verification.

[0086] Steps of residual analysis: draw residual graph, observe whether the residuals are randomly distributed, check whether the model assumptions (such as linear assumptions, homoscedasticity, etc.) are valid; test whether the residuals obey normal distribution, normality test.

[0087] The training is completed through verification, which ensures that the trained multivariate regression error calculation model has good predictive ability and stability for new data, and improves the generalization ability of the multivariate regression error calculation model.

[0088] 205. Obtain a trained multivariate regression error calculation model according to the regression coefficients.

[0089] For this embodiment, after obtaining the regression coefficient of the target, a multivariate regression error calculation model can be constructed:

[0090]

[0091] in, is the regression coefficient of the target. The multivariate regression error calculation model quantifies the impact of each parameter item on the electric energy metering error.

[0092] 206. Obtain a real-time parameter value corresponding to the parameter item, use the multivariate regression error calculation model to perform error prediction on the real-time parameter value to obtain an error value, obtain an electric energy metering prediction value, and obtain an electric energy metering correction value according to the error value and the electric energy metering prediction value.

[0093] For this embodiment, the real-time parameter value is Will Substitute into x1, Substituting into x2, ... Substitute into x p , get the error value of a source

[0094] As another implementation, real-time data values ​​are obtained from multiple data sources, and the error value of each source is calculated respectively. Then use the preset weights to fuse these error values The final error value Error is obtained to improve the accuracy and reliability of the error value.

[0095] The predicted value of electric energy metering is obtained directly, which is E original , electric energy measurement correction value E corrected =E original -Error.

[0096] 207. Update the covariance matrix using the real-time parameter value to obtain an updated covariance matrix.

[0097]

[0098] Among them, x new is (1, real-time parameter value), the real-time parameter value is P is the covariance matrix, P 更新 To update the covariance matrix.

[0099] 208. Update the regression coefficient using the updated covariance matrix, the real-time parameter value, and the actual value of electric energy measurement to obtain an updated regression coefficient, and determine an updated multivariate regression error calculation model according to the updated regression coefficient.

[0100] For this embodiment, the updated regression coefficient is updated using the updated covariance matrix, the real-time parameter value, and the actual value of electric energy metering to obtain an updated regression coefficient, including: calculating an incremental update step using the updated covariance matrix and the real-time parameter value; obtaining the actual value of electric energy metering, obtaining a prediction error based on the actual value of electric energy metering and the predicted value of electric energy metering, and updating the regression coefficient based on the prediction error and the incremental update step to obtain an updated regression coefficient.

[0101] Among them, the incremental update step size K is:

[0102]

[0103] Electric energy measurement actual value y new , with an error value from one source As the energy metering prediction value, the prediction error ∈ new :

[0104]

[0105]

[0106] in, for

[0107]

[0108] For this embodiment, as an implementation method, a trigger condition for model update is set, for example, when the prediction error exceeds a preset threshold, the model update process is triggered. Through the above-mentioned incremental learning algorithm, the regression coefficient can be gradually and real-time adjusted according to the new data without completely retraining the model, so as to achieve dynamic adaptation to environmental changes and adaptive adjustment of the model. As another implementation method, new data is collected regularly and used as a new training set with historical data. The frequency of regular updates is set, for example, a comprehensive model training is performed using a new training set every month or quarter. Through regular training, it is ensured that the model is always in the optimal state and can adapt to the changing environment and working conditions.

[0109] The present invention provides an ultra-high voltage direct current transmission electric energy metering method, device and equipment, which are applied to edge computing units. Through the technical solution of the present invention, edge computing units are deployed at key nodes, and the computing power of the edge computing units is used to perform error compensation and data processing tasks in real time, reduce data transmission delays, and improve system response speed. Multiple parameter items that affect electric energy metering are obtained, covering all key parameters of the ultra-high voltage transmission system, ensuring the comprehensiveness and accuracy of the data. By training the initial multivariate regression error calculation model, an accurate relationship model between each parameter item and the error of electric energy metering, namely, a multivariate regression error calculation model, is established to ensure the accuracy and stability of the model. The process of using the multivariate regression error calculation model to predict the error of real-time parameter values ​​corresponding to multiple parameter items is to use the regression coefficient as the weighting coefficient, and use the regression coefficient to fuse the multiple parameter items. The fused data is the error value. The error value is used for correction to ensure the high accuracy and real-time performance of the electric energy metering correction value. The regression coefficients are updated in real time using new data (the new data include the real-time parameter values ​​corresponding to the parameter items and the actual values ​​of the electric energy metering), thereby determining the updated multivariate regression error calculation model, that is, realizing adaptive adjustment, ensuring that the model always maintains high accuracy and high reliability in different environments and working conditions.

[0110] Further, as Figure 1 and Figure 2 The specific implementation of the method shown in the figure, the embodiment of the present invention provides a UHV DC transmission power metering device, such as Figure 3 As shown, the device includes: a determination module 31, a training module 32, and a correction module 33;

[0111] A determination module 31, used to determine a plurality of parameter items for electric energy metering, and obtain historical parameter values ​​corresponding to the parameter items;

[0112] A training module 32 is used to obtain the historical actual value of electric energy metering, train an initial multiple regression error calculation model according to the historical parameter value and the historical actual value of electric energy metering, and obtain a trained multiple regression error calculation model;

[0113] The correction module 33 is used to obtain the real-time parameter value corresponding to the parameter item, use the multivariate regression error calculation model to perform error prediction on the real-time parameter value to obtain the error value, obtain the electric energy metering prediction value, and obtain the electric energy metering correction value according to the error value and the electric energy metering prediction value.

[0114] In specific application scenarios, such as Figure 4As shown, a UHV DC transmission electric energy metering device also includes: an updating module 34, which can be specifically used to update the covariance matrix using the real-time parameter value to obtain an updated covariance matrix; update the regression coefficient using the updated covariance matrix, the real-time parameter value, and the actual value of electric energy measurement to obtain an updated regression coefficient; and determine an updated multivariate regression error calculation model based on the updated regression coefficient.

[0115] Correspondingly, in order to update the regression coefficient using the updated covariance matrix, the real-time parameter value, and the actual value of electric energy metering to obtain an updated regression coefficient, the update module 34 can be specifically used to calculate the incremental update step using the updated covariance matrix and the real-time parameter value; obtain the actual value of electric energy metering, obtain the prediction error based on the actual value of electric energy metering and the predicted value of electric energy metering, and update the regression coefficient based on the prediction error and the incremental update step to obtain an updated regression coefficient.

[0116] Correspondingly, in order to determine multiple parameter items for electric energy metering, the determination module 31 can be specifically used to obtain a first preset number of original parameter items for electric energy metering and historical original parameter values ​​corresponding to the original parameter items; preprocess the historical original parameter values ​​to obtain preprocessed historical original parameter values; obtain the historical actual values ​​of electric energy metering and the corresponding historical predicted values ​​of electric energy metering, and calculate the electric energy metering error based on the historical actual values ​​of electric energy metering and the historical predicted values ​​of electric energy metering; calculate the correlation between the preprocessed historical original parameter values ​​and the electric energy metering error, sort the correlation, and determine the first second preset number of original parameter items as parameter items for electric energy metering.

[0117] In specific application scenarios, such as Figure 4 As shown, a UHV DC transmission electric energy metering device also includes: a construction module 35, which can be specifically used to construct regression coefficient terms; construct 1 and the product between the historical parameter values ​​and the corresponding regression coefficient terms, add the products to obtain a model prediction value; construct the difference between the model prediction value and the actual historical value of the electric energy metering, and use the square of the difference as an initial multiple regression error calculation model.

[0118] Correspondingly, in order to train the initial multiple regression error calculation model according to the historical parameter values ​​and the actual historical values ​​of the electric energy metering to obtain a trained multiple regression error calculation model, the training module 32 can be specifically used to calculate the initial regression coefficient of each regression coefficient item when the initial multiple regression error calculation model is minimized according to the historical parameter values ​​and the actual historical values ​​of the electric energy metering; test the initial regression coefficient, and if the test passes, determine the initial regression coefficient as the target regression coefficient; if the test fails, continue training until the test passes; and obtain a trained multiple regression error calculation model according to the regression coefficient.

[0119] Accordingly, in order to test the initial regression coefficient, the training module 32 may be specifically configured to test the initial regression coefficient using K-fold cross validation and residual analysis.

[0120] It should be noted that for other corresponding descriptions of the functional units involved in the UHV DC transmission energy metering device provided in this embodiment, reference can be made to Figure 1 to Figure 2 The corresponding description will not be repeated here.

[0121] Based on the above Figure 1 to Figure 2 The method shown in the embodiment, accordingly, also provides a storage medium, which can be volatile or non-volatile, and stores a computer program, which is executed by a processor to implement the above-mentioned Figure 1 to Figure 2 The UHV DC transmission energy metering method shown.

[0122] Based on this understanding, the technical solution of the present invention can be embodied in the form of a software product, which can be stored in a storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of various implementation scenarios of the present invention.

[0123] Based on the above Figure 1 to Figure 2 The method shown and Figure 3 , Figure 4 In order to achieve the above-mentioned purpose, the present embodiment further provides a computer device, which includes a storage medium and a processor; the storage medium is used to store a computer program; the processor is used to execute the computer program to achieve the above-mentioned Figure 1 to Figure 2 The UHV DC transmission energy metering method shown.

[0124] Optionally, the computer device may further include a user interface, a network interface, a camera, a radio frequency (RF) circuit, a sensor, an audio circuit, a WI-FI module, etc. The user interface may include a display, an input unit such as a keyboard, etc., and the optional user interface may also include a USB interface, a card reader interface, etc. The network interface may optionally include a standard wired interface, a wireless interface (such as a WI-FI interface), etc.

[0125] Those skilled in the art will appreciate that the computer device structure provided in this embodiment does not constitute a limitation on the physical device, and may include more or fewer components, or a combination of certain components, or different component arrangements.

[0126] The storage medium may also include an operating network communication module. The operating is a program that manages the hardware and software resources of the above-mentioned computer device, supporting the operation of the information processing program and other software and / or programs. The network communication module is used to communicate between the components inside the storage medium, and to communicate with other hardware and software in the information processing entity device.

[0127] Through the description of the above implementation modes, those skilled in the art can clearly understand that the present invention can be implemented by means of software plus a necessary general hardware platform, or by hardware.

[0128] The present invention provides an ultra-high voltage direct current transmission electric energy metering method, device and equipment, which are applied to an edge computing unit. First, multiple parameter items for electric energy metering are determined, and historical parameter values ​​corresponding to the parameter items are obtained; then, the actual historical values ​​of electric energy metering are obtained, and an initial multivariate regression error calculation model is trained according to the historical parameter values ​​and the actual historical values ​​of electric energy metering to obtain a trained multivariate regression error calculation model; finally, the real-time parameter values ​​corresponding to the parameter items are obtained, and the real-time parameter values ​​are predicted by using the multivariate regression error calculation model to obtain error values, and electric energy metering prediction values ​​are obtained, and electric energy metering correction values ​​are obtained according to the error values ​​and the electric energy metering prediction values. Through the technical solution of the present invention, edge computing units are deployed at key nodes, and the computing power of the edge computing units is used to perform error compensation and data processing tasks in real time, reduce data transmission delays, and improve system response speed. Multiple parameter items that affect electric energy metering are obtained, covering all key parameters of the ultra-high voltage transmission system, ensuring the comprehensiveness and accuracy of the data. By training the initial multivariate regression error calculation model, an accurate relationship model between each parameter item and the error of electric energy metering, namely, the multivariate regression error calculation model, is established to ensure the accuracy and stability of the model. The process of using the multivariate regression error calculation model to predict the error of the real-time parameter values ​​corresponding to multiple parameter items is to use the regression coefficient as the weighting coefficient, and use the regression coefficient to fuse the multiple parameter items. The fused data is the error value. This error value is used for correction to ensure the high accuracy and real-time performance of the electric energy metering correction value.

[0129] Those skilled in the art will appreciate that the accompanying drawings are only schematic diagrams of a preferred implementation scenario, and the modules or processes in the accompanying drawings are not necessarily necessary for implementing the present invention. Those skilled in the art will appreciate that the modules in the devices in the implementation scenario can be distributed in the devices of the implementation scenario according to the description of the implementation scenario, or can be changed accordingly and located in one or more devices different from the present implementation scenario. The modules of the above-mentioned implementation scenario can be combined into one module, or can be further split into multiple submodules.

[0130] The above serial numbers of the present invention are only for description and do not represent the advantages and disadvantages of the implementation scenarios. The above disclosure is only a few specific implementation scenarios of the present invention, but the present invention is not limited thereto, and any changes that can be thought of by those skilled in the art should fall within the protection scope of the present invention.

Claims

1. A method for measuring electric energy of ultra-high voltage direct current transmission, characterized in that: Applied to an edge computing unit, the method comprises: Determine a plurality of parameter items for electric energy metering, and obtain historical parameter values ​​corresponding to the parameter items; Acquire historical actual values ​​of electric energy metering, train an initial multiple regression error calculation model according to the historical parameter values ​​and the historical actual values ​​of electric energy metering, and obtain a trained multiple regression error calculation model; Obtain the real-time parameter value corresponding to the parameter item, use the multivariate regression error calculation model to perform error prediction on the real-time parameter value to obtain the error value, obtain the electric energy metering prediction value, and obtain the electric energy metering correction value according to the error value and the electric energy metering prediction value.

2. The method according to claim 1, characterized in that The method further comprises: Using the real-time parameter value to update the covariance matrix, to obtain an updated covariance matrix; The regression coefficient is updated by using the updated covariance matrix, the real-time parameter value, and the actual value of electric energy measurement to obtain an updated regression coefficient; An updated multiple regression error calculation model is determined according to the updated regression coefficients.

3. The method according to claim 2, characterized in that The updating of the regression coefficient by using the updated covariance matrix, the real-time parameter value, and the actual value of the electric energy measurement to obtain the updated regression coefficient includes: Calculating an incremental update step size using the updated covariance matrix and the real-time parameter value; The actual value of electric energy measurement is obtained, a prediction error is obtained according to the actual value of electric energy measurement and the predicted value of electric energy measurement, and a regression coefficient is updated according to the prediction error and the incremental update step to obtain an updated regression coefficient.

4. The method according to claim 1, characterized in that: The determining of multiple parameter items for electric energy metering includes: Acquire a first preset number of original parameter items for electric energy metering and historical original parameter values ​​corresponding to the original parameter items; Preprocessing the historical original parameter values ​​to obtain preprocessed historical original parameter values; Obtaining a historical actual value of electric energy measurement and a corresponding historical predicted value of electric energy measurement, and calculating an electric energy measurement error according to the historical actual value of electric energy measurement and the historical predicted value of electric energy measurement; The correlation between the preprocessed historical original parameter value and the electric energy metering error is calculated, the correlation is sorted, and the first second preset number of original parameter items are determined as parameter items for electric energy metering.

5. The method according to claim 1, characterized in that Before training the initial multiple regression error calculation model according to the historical parameter values ​​and the actual historical values ​​of the electric energy metering to obtain the trained multiple regression error calculation model, the method further includes: Construct regression coefficient terms; Construct 1 and the product of the historical parameter value and the corresponding regression coefficient term, add the products to obtain the model prediction value; The difference between the model prediction value and the actual historical value of the electric energy metering is constructed, and the square of the difference is used as an initial multiple regression error calculation model.

6. The method according to claim 5, characterized in that The initial multiple regression error calculation model is trained according to the historical parameter values ​​and the actual historical values ​​of the electric energy metering to obtain a trained multiple regression error calculation model, including: The initial regression coefficient of each regression coefficient term is calculated based on the historical parameter value and the actual historical value of the electric energy metering so that the initial multiple regression error calculation model is minimized; Testing the initial regression coefficient, if the test passes, determining the initial regression coefficient as the regression coefficient of the target, if the test fails, continuing training until the test passes; A trained multivariate regression error calculation model is obtained according to the regression coefficients.

7. The method according to claim 6, characterized in that The testing of the initial regression coefficient comprises: The initial regression coefficients were tested using K-fold cross validation and residual analysis.

8. A UHV DC power transmission energy metering device, characterized in that: Applied to an edge computing unit, the device comprises: A determination module, used to determine a plurality of parameter items for electric energy metering, and obtain historical parameter values ​​corresponding to the parameter items; A training module is used to obtain the historical actual values ​​of electric energy metering, train an initial multiple regression error calculation model according to the historical parameter values ​​and the historical actual values ​​of electric energy metering, and obtain a trained multiple regression error calculation model; The correction module is used to obtain the real-time parameter value corresponding to the parameter item, use the multivariate regression error calculation model to perform error prediction on the real-time parameter value to obtain the error value, obtain the electric energy metering prediction value, and obtain the electric energy metering correction value according to the error value and the electric energy metering prediction value.

9. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the ultra-high voltage direct current transmission electric energy metering method as claimed in any one of claims 1 to 7 is implemented.

10. A computer device comprising a storage medium, a processor, and a computer program stored in the storage medium and executable on the processor, characterized in that: When the processor executes the computer program, the UHV DC transmission electric energy metering method according to any one of claims 1 to 7 is implemented.

Images