A method and system for correcting energy metering errors in electric vehicle charging piles

Abstract

This invention discloses a method and system for correcting energy metering errors in electric vehicle charging piles. The method includes collecting energy metering error data of charging piles under three-dimensional operating conditions (different power, temperature, and humidity) and constructing a three-dimensional operating condition error database; hierarchically filtering high-confidence neighbor points to be interpolated to obtain a standardized set of high-confidence neighbor points; calculating the radial distance between the standardized point to be interpolated and each standardized high-confidence neighbor point in the standardized set to optimize the interpolation width parameter of the radial basis function and dynamically selecting the basis function type; introducing error confidence to correct the weights of the radial basis function; and weighted summing the corrected radial basis function weights and the energy metering errors of the high-confidence neighbor point set to obtain the corrected energy metering error value for the point to be interpolated. This invention aims to achieve accurate and real-time energy metering error correction under all operating conditions to improve the reliability of charging pile metering.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of electric energy metering of charging piles, and particularly relates to an electric energy metering error correction method and system for electric vehicle charging piles. BACKGROUND

[0002] With the rapid development of the electric vehicle industry, as a key infrastructure for energy supply, the electric energy metering accuracy of charging piles is directly related to user rights, operator benefits and the fairness of power grid settlement. However, current charging piles are mostly deployed in outdoor environments with large fluctuations in temperature and humidity. Moreover, electric vehicle charging has formed a diversified pattern of "slow charging for energy supply, fast charging for emergency, and super charging for speed", and the power difference under different charging modes can be up to tens of times. Such multi-dimensional working condition characteristics make the electric energy metering process face complex interference. Temperature directly affects the sampling accuracy of the metering chip and the resistance characteristics of the manganese-copper shunt, humidity easily leads to the decline of the insulation performance of the voltage dividing resistor, and the current load fluctuation under different charging powers further aggravates the coupling interference of temperature and humidity on the metering elements. For example, in the super charging scenario, the local high temperature zone formed by the superposition of joule heat generated by large current and environmental temperature will cause nonlinear deviation of metering error; in the low temperature and humidity environment, small current sampling of slow charging is more sensitive to the slight error caused by temperature and humidity. The multi-dimensional dynamic coupling of temperature, humidity and charging power finally makes the error source of the charging pile metering system complex, and the error law presents significant nonlinear and time-varying characteristics, which has become a core problem to be solved in the industry.

[0003] The existing correction method for charging pile electric energy metering error first faces the limitations of incomplete core variable coverage and insufficient analysis of multi-parameter coupling mechanism. The current correction scheme generally has the defect of focusing on a single factor. Although some schemes have included two-dimensional environmental parameters such as temperature and humidity, they have failed to anchor the core differentiating variable of the charging scenario, namely the charging power. The huge difference in power parameters in slow charging, fast charging, super charging and other modes will directly change the working load and energy conversion characteristics of the metering module. The absence of this key variable not only leaves a gap in the core influencing factor coverage of the correction model from the source, but also causes the subsequent error analysis to lose the scene-based core anchor, making it impossible to meet the high-precision metering requirements in high-frequency dynamic charging scenarios. In practical applications, due to the complexity and diversity of working conditions, it is impossible to perform full-coverage sampling on all working condition combinations. However, interpolation methods can complete the error values in un-sampled working conditions through mathematical fitting based on limited sample data, thereby achieving approximate calculation of the error of all working conditions. However, in terms of the technical path of implementing charging pile electric energy metering error correction, existing interpolation methods all have obvious adaptation defects. Linear interpolation, although simple and efficient, can only depict the linear relationship between parameters and is completely unable to cope with the non-linear error pattern under the coupling of temperature, humidity and power, and is prone to correction deviation when power fluctuates or temperature and humidity change; Kriging interpolation can quantify the uncertainty of interpolation through the semi-variogram function, but it requires high uniformity of sample distribution, and the actual working conditions of charging piles have the characteristics of sparse extreme scenario samples and dense normal scenario samples, which will greatly increase the computational complexity and significantly reduce the accuracy in sparse sample areas; the nearest neighbor interpolation relies on the local correlation of sample points, and when there are abnormal samples caused by sensor drift, it is easy to directly pass the error to the correction result, with poor stability. Radial interpolation has strong fitting ability for non-linear data and can quickly calculate error compensation values in small samples or local working conditions through the local characteristics of radial basis functions, to some extent adapting to the continuous variation characteristics of some parameters in the charging pile metering scenario. However, existing radial interpolation methods rely on manual setting of key parameters, and improper parameter selection can lead to excessive dispersion of weight distribution or amplification of noise, and the fixed basis function type cannot adapt to the diversity of sample distribution, resulting in significant interpolation result deviation in scenarios such as power mutation and drastic temperature and humidity fluctuations; at the same time, the traditional correction model does not consider the transmission characteristics of errors between different working conditions, and lacks quantitative evaluation of the reliability of sample errors themselves. When some samples have large errors due to sensor drift, directly participating in interpolation will distort the correction result and make it difficult to meet the metering accuracy requirements throughout the life cycle. Therefore, how to achieve accurate and real-time electric energy metering error correction in all working conditions to improve the reliability of charging pile metering has become a key technical problem to be solved. SUMMARY

[0004] The technical problems solved by the present application are: In view of the above problems of the prior art, the present application provides an electric vehicle charging pile electric energy metering error correction method and system, and aims to realize accurate and real-time electric energy metering error correction under all working conditions to improve the reliability of charging pile metering.

[0005] To solve the above technical problems, the technical scheme adopted by the present application is:

[0006] An electric vehicle charging pile electric energy metering error correction method comprises the following steps:

[0007] S101, collecting charging pile electric energy metering error data under different power, temperature and humidity three-dimensional working conditions and constructing a three-dimensional working condition error database, wherein the charging pile electric energy metering error data comprises error mean and error standard deviation;

[0008] S102, selecting a high-impact high-confidence neighboring point set from the three-dimensional working condition error database; The high-impact high-confidence neighboring point set is composed of high-impact high-confidence neighboring points filtered in layers in the three-dimensional working condition error database The high-impact high-confidence neighboring point set is composed of high-impact high-confidence neighboring points filtered in layers in the three-dimensional working condition error database ;

[0009] S103, normalizing the to-be-interpolated point , calculating the radial distance between each normalized high-confidence neighboring point in the high-confidence neighboring point normalized set and the normalized to-be-interpolated point , and extracting the maximum and minimum values of the radial distance;

[0010] S104, combining the maximum and minimum values of the radial distance and the size of the high-confidence neighboring point normalized set to calculate the interpolation width parameter of the radial basis function; dynamically selecting the basis function type according to the size of the high-confidence neighboring point normalized set and the interpolation width parameter of the radial basis function;

[0011] S105, introducing error confidence to correct the weight of the radial basis function, weighting the electric energy metering error of the high-confidence neighboring point set and the electric energy metering error correction value of the to-be-interpolated point .

[0012] Optionally, in step S102, the high-impact high-confidence neighboring point set is composed of high-impact high-confidence neighboring points filtered in layers in the three-dimensional working condition error database

[0013] ​​​S201 refers to the set of sample points extracted from the three-dimensional working condition error database. For each sample point in the dataset, calculate the sample point and the set of sample points. The Euclidean distances to other sample points are calculated, and a structure with size is constructed. Distance matrix , For the set of sample points Size; interpolation point Combining preset temperature and humidity ranges, in the sample point set Select the interpolation points from the middle The set of neighboring sample points ;

[0014] S202, from the distance matrix Extract all A non-diagonal element, After sorting the non-diagonal elements in ascending order, select the preset first element from the sorted sequence. The Euclidean distance of the two digits is used as the cutoff distance. ;

[0015] S203 refers to the set of sample points extracted from the three-dimensional working condition error database. For each sample point in the data, based on the cutoff distance... interpolation point The set of neighboring sample points Calculate the set of sample points Local density of each sample point in the sample;

[0016] S204, Based on the local density of each sample point and a preset density threshold, select sample points whose local density is greater than the preset density threshold and add them to the high confidence point set. ;

[0017] S205, Take the set of high confidence points and neighboring sample point set The intersection of these points forms the high-confidence nearest neighbor set. ; Calculate the set of high-confidence nearest neighbors The charging pile power metering error data, as well as the mean and standard deviation of power, temperature, and humidity, are used to analyze the high-confidence nearest neighbor set. Standardization to obtain a standardized set of high-confidence nearest neighbors .

[0018] Optionally, in step S203, based on the cutoff distance interpolation point The set of neighboring sample points Calculate the set of sample points The function expression of the local density of each sample point is:

[0019] ;

[0020] wherein, is the local density of the sample point , the sample point is the sample point corresponding to the i-th temperature, the j-th humidity and the k-th power in the sample point set , , , , , and are the gradient indexes of the power, the temperature and the humidity respectively, is the neighboring sample point set of the point to be interpolated , is the truncation distance, is the Euclidean distance between the sample point and the neighboring sample point , the neighboring sample point is the sample point in , , and are the gradient indexes of the power, the temperature and the humidity respectively.

[0021] Optionally, the function expressions for calculating the mean and the standard deviation of the three dimensions of power, temperature and humidity in step S205 are as follows:

[0022] , ;

[0023] , ;

[0024] , ;

[0025] wherein, and are the mean and the standard deviation of the temperature dimension respectively, is the number of gradient indexes of the temperature, and are the mean and the standard deviation of the humidity dimension respectively, is the number of gradient indexes of the humidity, and are the mean and the standard deviation of the power dimension respectively, is the number of gradient indexes of the power, , and are the high-confidence neighboring point sets respectively Medium to high confidence neighboring points Temperature, humidity and power in the middle, , For the set of high confidence neighbor points and the standardized set of high-confidence nearest neighbors Size; in step S205, the mean and standard deviation are combined for the high-confidence nearest neighbor set. The standardized function expression is as follows:

[0026] , , ;

[0027] in, , and These are the standardized sets of high-confidence nearest neighbors. Medium-standardized high-confidence neighbor points Temperature, humidity, and power within.

[0028] Optionally, in step S103, the standardized interpolation points are calculated. High-confidence neighbor set The functional expression for the radial distance between each standardized high-confidence neighbor is:

[0029] ;

[0030] in, The standardized interpolation points Standardized high-confidence neighbor points radial distance between , and These are the standardized interpolation points. Temperature, humidity and power in the middle, , and These are the standardized high-confidence nearest neighbors. Temperature, humidity, and power within.

[0031] Optionally, in step S104, the maximum and minimum radial distances are combined with the standardized set of high-confidence nearest neighbors. The function expression for the interpolation width parameter of the radial basis function, calculated from its magnitude, is as follows:

[0032] ;

[0033] in, The interpolation width parameter for the radial basis functions. and These are the maximum and minimum values ​​of the radial distance, respectively. For the set of high confidence neighbor points and the standardized set of high-confidence nearest neighbors Size; based on the high-confidence nearest neighbor set The function expression for dynamically selecting the basis function type using the parameters of size and interpolation width of the radial basis functions is:

[0034] ;

[0035] in, For the selected basis functions, The interpolation width parameter for the radial basis functions. The standardized interpolation points Standardized high-confidence neighbor points radial distance between For the set of high confidence neighbor points and the standardized set of high-confidence nearest neighbors Size.

[0036] Optionally, when introducing the error confidence level to correct the weights of the radial basis function in step S105, the calculation function expression for the weights of the corrected radial basis function is as follows:

[0037] ;

[0038] ;

[0039] in, For the set of high confidence neighbor points Medium to high confidence neighboring points The corresponding weights of the modified radial basis function, For the error confidence level, For the selected basis functions, , and These are the standardized high-confidence nearest neighbors. Temperature, humidity and power in the middle, For the high-confidence neighbor set, For the set of high confidence neighbor points Medium to high confidence neighboring points standard deviation The preset maximum error for the scenario; the weights of the corrected radial basis function and the set of high-confidence neighbor points. The interpolation point is obtained by weighted summation of the power metering errors. The functional expression for the power metering error correction value is:

[0040] ;

[0041] in, Points to be interpolated The power metering error correction value, For the set of high confidence neighbor points and the standardized set of high-confidence nearest neighbors Size, For the set of high confidence neighbor points Medium to high confidence neighboring points The corresponding weights of the modified radial basis function, For the set of high confidence neighbor points The error in electricity metering.

[0042] The present invention also provides an electric vehicle charging pile power metering error correction system, including a microprocessor and a memory connected to each other, wherein the microprocessor is programmed or configured to execute the electric vehicle charging pile power metering error correction method.

[0043] The present invention also provides a computer-readable storage medium storing a computer program or instructions that are programmed or configured to execute the electric vehicle charging pile energy metering error correction method by a processor.

[0044] The present invention also provides a computer program product, including a computer program or instructions, which are programmed or configured to execute the electric vehicle charging pile energy metering error correction method by a processor.

[0045] Compared with existing technologies, the present invention can achieve the following beneficial effects: Addressing the problems of existing technologies that only consider the two-dimensional influence of temperature and humidity, fail to adapt to diverse charging power differences, and ignore multi-parameter coupling mechanisms, the present invention constructs a three-dimensional error database of "temperature-humidity-power," incorporating different power conditions such as slow charging, fast charging, and supercharging into the core influence dimension. It systematically quantifies the influence of the cross-coupling effect of these three factors on measurement errors, breaking through the bottleneck of traditional single-factor or two-dimensional correction due to dimensional deficiencies. This allows for accurate capture of the nonlinear and time-varying characteristics of errors under dynamic multi-parameter interaction, fundamentally solving the problem of correction deviation and comprehensively covering the measurement error correction needs of diverse charging scenarios. To address the problems of existing radial interpolation algorithms that rely on manually set parameters, have fixed basis functions that cannot adapt to sample diversity, and are prone to interpolation deviations in scenarios with sudden power changes and drastic temperature and humidity fluctuations, this invention innovatively optimizes the radial basis function interpolation algorithm. It adaptively calculates the width parameter λ based on the number and distance characteristics of high-confidence neighbor points and dynamically selects the basis function type according to the number of samples. Simultaneously, it introduces an error confidence correction factor and calibrates the interpolation weights based on sample error reliability, avoiding problems such as excessive weight dispersion or noise amplification caused by improper manual parameters. This significantly improves the adaptability of the interpolation algorithm to dynamic operating conditions and solves the problem of insufficient accuracy in charging pile power metering interpolation algorithms. This invention addresses the challenges of existing energy metering error correction technologies that only consider the two-dimensional effects of temperature, humidity, and power, and integrate multi-dimensional parameter coupling mechanisms with adaptive interpolation optimization. Firstly, it overcomes the limitations of existing technologies that only consider the two-dimensional effects of temperature and humidity. These technologies fail to adapt to the diverse charging power differences of slow charging, fast charging, and supercharging, and neglect significant power differences under multi-parameter coupling mechanisms, making it difficult to meet the high-precision metering requirements of high-frequency dynamic charging scenarios. Secondly, it solves the problems in existing energy metering error correction technologies, such as radial interpolation algorithms relying on manually set parameters and fixed basis functions, which cannot adapt to sample diversity. This leads to interpolation deviations under scenarios of sudden power changes and drastic temperature and humidity fluctuations, and insufficient accuracy under dynamic operating conditions. The invention achieves accurate and real-time energy metering error correction under all operating conditions, improving the reliability of charging pile metering. Attached Figure Description

[0046] Figure 1 This is a schematic diagram of the basic process of the method in an embodiment of the present invention.

[0047] Figure 2 This is a schematic diagram illustrating the principle of radial basis function interpolation in an embodiment of the present invention. Detailed Implementation

[0048] To enable those skilled in the art to better understand the technical solutions of the present invention, the technical solutions of the present invention will be further described in detail below with reference to the accompanying drawings in the embodiments of the present invention.

[0049] like Figure 1As shown, the method for correcting the energy metering error of electric vehicle charging piles in this embodiment includes the following steps:

[0050] S101, Collect charging pile power metering error data under three-dimensional operating conditions of different power, temperature and humidity and construct a three-dimensional operating condition error database. The charging pile power metering error data includes the error mean and error standard deviation.

[0051] S102, for interpolation points In the three-dimensional working condition error database, high-confidence nearest neighbor points with high influence are selected hierarchically to form a high-confidence nearest neighbor set. Set the high-confidence nearest neighbors Standardization yields the standardized set of high-confidence nearest neighbors. ;

[0052] S103, the interpolation point Standardization processing, calculating the standardized interpolation points. High-confidence neighbor set The radial distance between each standardized high-confidence neighbor point is calculated, and the maximum and minimum values ​​of the radial distance are extracted.

[0053] S104, combining the maximum and minimum radial distances with the standardized set of high-confidence nearest neighbors. The interpolation width parameter of the radial basis function is obtained by calculating its magnitude; based on the high-confidence nearest neighbor set... The size of the radial basis function and the interpolation width parameter dynamically select the basis function type;

[0054] S105 introduces error confidence correction to the weights of the radial basis function, and combines the corrected weights of the radial basis function with the set of high-confidence nearest neighbors. The interpolation point is obtained by weighted summation of the power metering errors. The power metering error correction value.

[0055] In this embodiment, S101, when collecting charging pile power metering error data under different power, temperature, and humidity three-dimensional operating conditions and constructing a three-dimensional operating condition error database, the configuration of temperature and humidity test conditions includes: using a constant temperature and humidity chamber to discretize the temperature and humidity ranges covered by the system at preset intervals. After discretization, the temperature range is set to N different temperature ranges, and after discretization, the humidity range has M different humidity levels. A = M × N sets of temperature and humidity test conditions are used. The configuration of power test conditions includes: to simulate charging piles with different power levels such as slow charging, fast charging, and supercharging, the charging power is set to three different power conditions: slow charging power, fast charging power, and supercharging power. The number of power test conditions for the three different power conditions is recorded as L. The collected error data is stored according to the three-dimensional coordinates of temperature, humidity, and power, with a total of H three-dimensional samples, H = A × L. A three-dimensional error database is constructed, where each sample point... ,in ; ; , , , The indexes are for power, temperature, and humidity gradients. Specifically, this embodiment utilizes a TH-800F constant temperature and humidity chamber, with a temperature control range of -40℃ to 80℃ and a humidity control range of 20%RH to 98%RH, depending on the environment at the testing site. The power simulator used is a Keysight N6705C DC power analyzer. Temperature gradients: -30℃, -20℃, -10℃, 0℃, 10℃, 20℃, 30℃, 40℃, 50℃, 60℃. Humidity gradients: 30%RH, 40%RH, 50%RH, 60%RH, 70%RH, 80%RH, 90%RH. Power gradients: slow charging 3.5kW, 7kW, 11kW; fast charging 60kW, 120kW, 150kW; supercharging 240kW, 360kW, 600kW. The standard meter used is the Wasion DTSD341-MC3, with active power accuracy reaching 0.5S level. The number of temperature gradients is N=10, the number of humidity gradients is M=7, the number of power conditions is L=9, the number of temperature and humidity combinations is A=M×N=70, and the total number of three-dimensional samples is H=70×9=630.

[0056] During the three-dimensional operating condition power metering error experiment, the temperature and humidity environmental chamber parameters were set to the target operating condition. The power source outputs the target power. The load needs to be charged. Hours The time taken is generally no more than 10 hours. The metering values ​​of the charging pile under test are recorded synchronously. Compared with standard meter measurement value ( ), calculate the first Calibration error of the test:

[0057] ;

[0058] Calculate the mean error and standard deviation of error :

[0059] ;

[0060] .

[0061] In charging pile metering error monitoring, interpolation calculations are required based on a three-dimensional sample space of power, temperature, and humidity. When the total number of samples reaches H, calculating the distance for all samples would lead to a surge in computational load, making it difficult to meet the real-time requirements of online monitoring. Simultaneously, some sample points, due to their sparse distribution (such as combinations of extreme power and extreme temperature and humidity), contribute minimally to the interpolation results, and blindly including them would introduce errors. Therefore, it is necessary to balance efficiency and accuracy by stratifying and selecting high-impact sample points. In step S102 of this embodiment, the interpolation points... In the three-dimensional working condition error database, high-confidence nearest neighbor points with high influence are selected hierarchically to form a high-confidence nearest neighbor set. include:

[0062] S201 refers to the set of sample points extracted from the three-dimensional working condition error database. For each sample point in the dataset, calculate the sample point and the set of sample points. The Euclidean distances to other sample points are calculated, and a structure with size is constructed. Distance matrix , For the set of sample points Size; interpolation point Combining preset temperature and humidity ranges, in the sample point set Select the interpolation points from the middle The set of neighboring sample points ; the sample points in the above three-dimensional error database A set is called a set. Ω, Right now For a certain point in three-dimensional space to be interpolated... ,if Through temperature range Humidity range The initial selection of the set of neighboring grid points is denoted as . satisfy:

[0063] ;

[0064] In this embodiment, , Furthermore, the temperature and humidity range must be within the experimental testing range, passing through the temperature range [ T q 20℃ T q +20℃, humidity range R q 20%, R q [+20%] Initially, a set of neighboring grid points was selected, denoted as... A total of 144 points;

[0065] S202, from the distance matrix Extract all A non-diagonal element, After sorting the non-diagonal elements in ascending order, select the preset first element from the sorted sequence. The Euclidean distance of the two digits is used as the cutoff distance. ;

[0066] For sets any point in the middle Calculate with other points , ， The Euclidean distance in the temperature and humidity space is:

[0067] ;

[0068] Build size is Distance matrix ,in:

[0069] and diagonal elements In this embodiment, from the distance matrix Extract all off-diagonal elements. The number of off-diagonal elements is:

[0070] ;

[0071] right B Sort the distance values ​​in ascending order, and take the nth distance value from the sequence. The distance value of the bit is used as the cutoff distance. ,in Determined by a 3% proportion of the sequence length. , The function to round down;

[0072] S203 refers to the set of sample points extracted from the three-dimensional working condition error database. For each sample point in the data, based on the cutoff distance... interpolation point The set of neighboring sample points Calculate the set of sample points Local density of each sample point in the sample;

[0073] S204, Based on the local density of each sample point and a preset density threshold, select sample points whose local density is greater than the preset density threshold and add them to the high confidence point set. ;

[0074] S205, Take the set of high confidence points and neighboring sample point set The intersection of these points forms the high-confidence nearest neighbor set. ; Calculate the set of high-confidence nearest neighbors The charging pile power metering error data, as well as the mean and standard deviation of power, temperature, and humidity, are used to analyze the high-confidence nearest neighbor set. Standardization to obtain a standardized set of high-confidence nearest neighbors .

[0075] Local density reflects the distribution density around the sample. In step S203 of this embodiment, it is based on the cutoff distance. interpolation point The set of neighboring sample points Calculate the set of sample points The functional expression for the local density of each sample point is:

[0076] ;

[0077] in, For sample points Local density, sample points For the set of sample points The Middle The temperature, the first The humidity and the first Each power corresponding to a sample point , and These are gradient indices for power, temperature, and humidity, respectively. Points to be interpolated The set of neighboring sample points, To cut off the distance, For sample points and neighboring sample points Euclidean distance, neighboring sample points for The sample points in , and These are gradient indices for power, temperature, and humidity, respectively. In step S204, based on the local density of each sample point and a preset density threshold, sample points with local densities greater than the preset density threshold are selected and added to the high-confidence point set. This includes calculating the local density mean of the aforementioned grid points:

[0078] ,

[0079] Set density threshold (Typical value is taken) =1.1 ).

[0080] Screening out local density The points are placed into the high confidence point set. .

[0081] In step S205, the set of high confidence points is taken. and neighboring sample point set The intersection of these points forms the high-confidence nearest neighbor set. , can be represented as: The element is ( ), The number of elements in the high-confidence neighbor set corresponds to the energy metering error. ; Calculate the set of high-confidence nearest neighbors The charging pile power metering error data, as well as the mean and standard deviation of power, temperature, and humidity, are used to analyze the high-confidence nearest neighbor set. Standardization to obtain a standardized set of high-confidence nearest neighbors .

[0082] In step S205, to eliminate the set of high-confidence neighbor points... The units and numerical ranges of power, temperature, and humidity differ, so each dimension is standardized, and the mean value of each dimension is calculated. and standard deviation The functional expressions for calculating the mean and standard deviation of the three dimensions of power, temperature, and humidity in step S205 are as follows:

[0083] , ;

[0084] , ;

[0085] , ;

[0086] in, and These are the mean and standard deviation for the temperature dimension, respectively. This represents the number of temperature gradient indices. and These are the mean and standard deviation for the humidity dimension, respectively. This represents the number of gradient indices for humidity. and These are the mean and standard deviation of the power dimension, respectively. This represents the number of gradient indices for the power. , and They are respectively the high confidence neighbor set Medium to high confidence neighboring points Temperature, humidity and power in the middle, , For the set of high confidence neighbor points and the standardized set of high-confidence nearest neighbors The size of the high-confidence nearest neighbor set. In step S205 of this embodiment, the mean and standard deviation are combined to evaluate the high-confidence nearest neighbor set. The standardized function expression is as follows:

[0087] , , ;

[0088] in, , and These are the standardized sets of high-confidence nearest neighbors. Medium-standardized high-confidence neighbor points Temperature, humidity and power in the middle, .

[0089] interpolation point The standardized points are the points to be interpolated. In step S103 of this embodiment, the standardized interpolation points are calculated. High-confidence neighbor set The functional expression for the radial distance between each standardized high-confidence neighbor is:

[0090] ;

[0091] in, The standardized interpolation points Standardized high-confidence neighbor points radial distance between , and These are the standardized interpolation points. Temperature, humidity and power in the middle, , and These are the standardized high-confidence nearest neighbors. Temperature, humidity, and power. Existing radial interpolation function width parameters. Relying on manual settings, if Excessive weight distribution leads to scattered distribution, and interference from distant neighboring samples in the interpolation results; if If the weights are too small, the concentration of power will be too high, amplifying the noise of nearest neighbor samples. Furthermore, the radial interpolation function uses a fixed single basis function (such as a Gaussian function), which cannot adapt to the diversity of charging pile sample distributions. The above formula is used to iterate through the set of high-confidence nearest neighbors. All of them Q Find the highest and lowest radial distances between the three high-confidence neighboring points.

[0092] ;

[0093] .

[0094] In step S104 of this embodiment, the maximum and minimum radial distances are combined with the standardized set of high-confidence nearest neighbors. The function expression for the interpolation width parameter of the radial basis function, calculated from its magnitude, is as follows:

[0095] ;

[0096] in, The interpolation width parameter for the radial basis functions. and These are the maximum and minimum values ​​of the radial distance, respectively. For the set of high confidence neighbor points and the standardized set of high-confidence nearest neighbors The size of the sample; the denser the sample, the better. The smaller the value, the more focused it is on nearest neighbor samples; the sparser the samples, the better. The larger.

[0097] In this embodiment, the high-confidence nearest neighbor set is used as the basis. The function expression for dynamically selecting the basis function type using the parameters of size and interpolation width of the radial basis functions is:

[0098] ;

[0099] in, For the selected basis functions, The interpolation width parameter for the radial basis functions. The standardized interpolation points Standardized high-confidence neighbor points radial distance between For the set of high confidence neighbor points and the standardized set of high-confidence nearest neighbors Size.

[0100] Traditional radial interpolation calculates weights based solely on distance, by default. A smaller value indicates a greater contribution of high-confidence neighboring points to the interpolation. However, in the multi-factor coupled scenario of charging piles, sample errors are significantly affected by the interaction of temperature, humidity, and power. If the distance weight continues to be relied upon, high-error samples will dominate the interpolation results, leading to a decrease in correction accuracy. Therefore, this embodiment introduces an error confidence level. ( Adjusting the weights using the sample's own error Maximum error with the scene The normalized exponential function dynamically calibrates sample weights, and the error confidence level. The definition of is:

[0101] ;

[0102] in, The maximum error for the preset scenario is defined as the maximum permissible error of the charging pile. Without loss of generality, the maximum permissible error for AC charging piles is set to 0.5%, and for DC charging piles, it is set to 1%. The error confidence level is incorporated into the radial basis function weights. In step S105, when the error confidence level is introduced to correct the weights of the radial basis function, the expression for the corrected weights of the radial basis function is as follows:

[0103] ;

[0104] in, For the set of high confidence neighbor points Medium to high confidence neighboring points The corresponding weights of the modified radial basis function, For the error confidence level, For the selected basis functions, , and These are the standardized high-confidence nearest neighbors. Temperature, humidity and power in the middle, For the high-confidence neighbor set, For the set of high confidence neighbor points Medium to high confidence neighboring points The standard deviation of the set. In Q A number of high-confidence neighbor samples, based on the unique indexes of these samples in the 3D database. kThe error confidence level calculated from the matched sample is directly extracted. As the error confidence level of high-confidence neighbor samples ; It is a radial basis function. It is the set to be interpolated Radial distance between elements. Molecular description. k Confidence level of a specific high-confidence neighbor point With radial basis functions The product of and reflects the contribution of a single point to the error correction; the denominator is a product of all . Q The confidence level of each high-confidence neighbor is calculated sequentially. With corresponding radial basis functions The summation of the products of all points reflects the total contribution of all points to the error correction.

[0105] Correction value for power metering error at the interpolation point Errors from high confidence neighbor points With corresponding weights The weighted sum is obtained. For example... Figure 2 As shown, in this embodiment, the weights of the modified radial basis function and the set of high-confidence neighbor points are... The interpolation point is obtained by weighted summation of the power metering errors. The functional expression for the power metering error correction value is:

[0106] ;

[0107] in, Points to be interpolated The power metering error correction value, For the set of high confidence neighbor points and the standardized set of high-confidence nearest neighbors Size, For the set of high confidence neighbor points Medium to high confidence neighboring points The corresponding weights of the modified radial basis function, For the set of high confidence neighbor points The method for correcting the metering error of charging piles is as follows: A high-confidence neighbor sample set is obtained from the original sample library. For the interpolation condition, the standardized three-dimensional distance of each sample is calculated. Combined with its error confidence, a radial basis function is dynamically selected and the interpolation weight is calculated. The average error of each sample is weighted and summed according to the corresponding weights using the above formula. The result is the corrected metering error value of the charging pile under the interpolation condition. This value comprehensively considers the reliability of sample errors and the similarity of operating conditions, and can accurately correct metering errors in multi-factor coupled scenarios. It is compatible with full-power mode, improves metering accuracy and robustness, and can be used for charging pile metering error correction.

[0108] To verify the electric vehicle charging pile power metering error correction method of this embodiment, a three-dimensional error database was constructed through a full factorial combination experiment. When the power was fixed at 60kW, the power error was tested under different temperature and humidity conditions to show the influence of temperature and humidity on the error. When the temperature and humidity were fixed at 25℃ and 50%, the error of charging piles with different power was tested. The power metering error correction results under different power, temperature and humidity conditions are shown in Tables 1 to 3.

[0109] Table 1: Effects of temperature and humidity on power error under a fixed power of 60kW and 1 hour of testing per operating condition.

[0110]

[0111] Table 2: Experimental data on the error of charging piles with different power ratings after a 1-hour test at a constant temperature and humidity of 25℃ and 50%.

[0112]

[0113] Table 3: Results of power metering error correction under different power, temperature and humidity conditions

[0114]

[0115] As shown in Tables 1 to 3, the measurement error fluctuates significantly in the multi-dimensional coupling scenario before correction. At a fixed power of 60kW, the error reaches -0.9012% at -30℃ and 90% humidity, 0.8998% at 40℃ and 40% humidity, and 0.8969% at 60℃ and 30% humidity. At a fixed temperature of 25℃ and 50% humidity, the error for 11kW slow charging is -0.8780%, and the error for 600kW supercharging is -0.4138%. After correction by the method of this invention, the error under all operating conditions is significantly reduced: the error under the 3.5kW slow charging (-28℃, 30%) condition was 0.2343% before correction, and reduced to 0.0171% after correction; the error under the 60kW fast charging (-5℃, 89%) condition was -0.5666% before correction, and reduced to only 0.012% after correction; the error under the 600kW supercharging (65℃, 79%) condition was 0.8510% before correction, and controlled within 0.1668% after correction. This fully verifies the accurate correction capability of the electric vehicle charging pile power metering error correction method of this embodiment for multi-dimensional coupled errors, as well as its adaptability to all power modes of slow charging, fast charging, and supercharging.

[0116] Furthermore, this embodiment also provides an electric vehicle charging pile energy metering error correction system, including a microprocessor and a memory interconnected. The microprocessor is programmed or configured to execute the electric vehicle charging pile energy metering error correction method. This system quantifies the influence of the interaction of temperature, humidity, and power on metering errors by constructing a three-dimensional error database. The core of the system lies in capturing global nonlinear trends and local dynamic details, innovatively optimizing the radial basis function interpolation algorithm, and adaptively calculating the width parameter. The system dynamically selects the basis function type based on the sample size, invents an error confidence correction factor, and combines it with an error propagation chain formed through online iteration. This ultimately achieves accurate and real-time correction of metering errors under all operating conditions, improving the accuracy and reliability of charging pile metering. This embodiment also provides a computer-readable storage medium storing a computer program or instructions programmed or configured to execute the electric vehicle charging pile energy metering error correction method via a processor. This embodiment further provides a computer program product, including a computer program or instructions programmed or configured to execute the electric vehicle charging pile energy metering error correction method via a processor.

[0117] Those skilled in the art will understand that the technical solutions provided by this invention may take the form of a method, system, or computer program product. Therefore, this invention may take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this invention may take the form of a computer program product embodied on one or more computer-readable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, produce an implementation of the flowchart... Figure 1 One or more processes and / or boxes Figure 1 The computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1The functions specified in one or more boxes. These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable apparatus for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0118] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. An electric vehicle charging pile electric energy metering error correction method, characterized in that, Comprising the following steps: S101, collecting charging pile electric energy measurement error data under different power, temperature and humidity three-dimensional working conditions and constructing a three-dimensional working condition error database, the charging pile electric energy measurement error data including error mean and error standard deviation; S102, to the interpolation point In the three-dimensional working condition error database, high-impact high-confidence neighboring points are hierarchically screened to form a high-confidence neighboring point set , the high-confidence neighboring point set is standardized to obtain a high-confidence neighboring point standardized set ; S103, obtaining a plurality of high-confidence neighboring points of the to-be-interpolated point S104, performing a normalization process on the to-be-interpolated point and the high-confidence neighboring points to obtain normalized to-be-interpolated points and normalized high-confidence neighboring points S105, obtaining a normalized high-confidence neighboring point set S106, calculating a radial distance between each normalized high-confidence neighboring point in the normalized high-confidence neighboring point set, and extracting a maximum value and a minimum value of the radial distance S104, combine the maximum and minimum values of the radial distance and the high-confidence neighboring point standardized set to calculate the interpolation width parameter of the radial basis function; according to the size of the high-confidence neighboring point standardized set to dynamically select the basis function type according to the size of the high-confidence neighboring point standardized set and the interpolation width parameter of the radial basis function; S105, introducing error confidence correction radial basis function weight, weighting and high confidence neighboring point set of the modified radial basis function weight The electric energy measurement error weighted sum of the to-be-interpolated point The electric energy measurement error correction value of the to-be-interpolated point The step S102 is to determine the point to be interpolated In the three-dimensional working condition error database, the high-impact and high-confidence neighboring point set is composed of the high-confidence neighboring points screened in layers Comprise: S201 refers to the set of sample points extracted from the three-dimensional working condition error database. For each sample point in the dataset, calculate the sample point and the set of sample points. The Euclidean distances to other sample points in the dataset are calculated, and a structure with a size of [size missing] is constructed. Distance matrix , For the set of sample points Size; interpolation point Combining preset temperature and humidity ranges, in the sample point set Select the interpolation points from the middle The set of neighboring sample points ; S202, extracting all non-diagonal elements from the distance matrix S203, sorting the non-diagonal elements in ascending order S204, selecting the preset Euclidean distance at the first position in the sorted sequence as the cutoff distance ;​​ S203, for each sample point in the sample point set extracted from the three-dimensional working condition error database, calculating a local density of the sample point according to a truncation distance, a neighboring sample point set of the sample point to be interpolated, and a sample point set of the neighboring sample point set. the sample point set​​​​ S204, according to the local density of each sample point, combine the preset density threshold to filter out the sample points with local density greater than the preset density threshold and put them into the high confidence point set ; S205, take the high confidence point set And the intersection of the adjacent sample point set Composed of high confidence adjacent point set ; Calculate the charging pile electric energy metering error data of the high confidence adjacent point set , and the mean and standard deviation of the three dimensions of power, temperature and humidity, combine the mean and standard deviation with the high confidence adjacent point set Standardization to obtain high confidence adjacent point standardized set .

2. The method for correcting the electric energy metering error of the electric vehicle charging pile according to claim 1, characterized in that, In step S203, the function expression of the local density of each sample point in the sample point set is calculated according to the truncation distance , the neighboring sample point set of the to-be-interpolated point The function expression of the local density of each sample point in the sample point set is calculated according to the truncation distance ​​ ； wherein, is the local density of the sample point , the sample point is the sample point in the sample point set , the th temperature, the th humidity and the th power, , and are the gradient indices of power, temperature and humidity, respectively, is the neighboring sample point set of the point to be interpolated, is the cut-off distance, is the Euclidean distance between the sample point and the neighboring sample point , the neighboring sample point is the sample point in , , and are the gradient indices of power, temperature and humidity, respectively.

3. The method for correcting the electric energy metering error of the electric vehicle charging pile according to claim 1, characterized in that, The function expression for calculating the mean and standard deviation of the three dimensions of power, temperature and humidity in step S205 is: ， ； ， ； ， ； wherein, and are the mean and standard deviation of the temperature dimension, respectively, is the number of gradient indices of the temperature, and are the mean and standard deviation of the humidity dimension, respectively, is the number of gradient indices of the humidity, and are the mean and standard deviation of the power dimension, respectively, is the number of gradient indices of the power, , and are the high-confidence neighboring points sets of temperature, humidity and power, respectively, in , , is the size of the high-confidence neighboring points set and the high-confidence neighboring points normalized set ; the function expression for normalizing the high-confidence neighboring points set in step S205 in combination with the mean and standard deviation is: ， ， ； wherein, , and are the normalized high-confidence neighboring points in the normalized high-confidence neighboring points temperature, humidity, and power in .

4. The method for correcting the electric energy metering error of the electric vehicle charging pile according to claim 1, characterized in that, The function expression of the radial distance between each standardized high-confidence neighboring point in the set of standardized high-confidence neighboring points is: a set of standardized high-confidence neighboring points The function expression of the radial distance between each standardized high-confidence neighboring point in the set of standardized high-confidence neighboring points is: ； wherein is the radial distance between the normalized point to be interpolated , the normalized high-confidence neighboring point , , and are the temperature, humidity and power in the normalized point to be interpolated , , and are the temperature, humidity and power in the normalized high-confidence neighboring point .

5. The method for correcting the electric energy metering error of the electric vehicle charging pile according to claim 1, characterized in that, Step S104 of combining the maximum and minimum values of the radial distance and the high-confidence neighboring point standardized set The function expression of the interpolation width parameter of the radial basis function calculated from the size of the high-confidence neighboring point standardized set is: ； wherein is an interpolation width parameter of the radial basis function, and are maximum and minimum values of the radial distance, respectively, is a size of the set of high-confidence neighboring points and a size of the set of high-confidence neighboring points normalized According to the size of the set of high-confidence neighboring points normalized and the interpolation width parameter of the radial basis function, a function expression of dynamically selecting the type of the basis function is: ； wherein is a selected basis function, is an interpolation width parameter of the radial basis function, is a normalized point to be interpolated , a normalized high-confidence neighbor point a radial distance between is a set of high-confidence neighbor points and a normalized set of high-confidence neighbor points a size of 6. The method for correcting the electric energy metering error of the electric vehicle charging pile according to claim 1, characterized in that, When the error confidence correction radial basis function weight is introduced in step S105, the calculation function expression of the corrected radial basis function weight is: ； ； wherein, is a high-confidence neighboring point set is a medium-confidence neighboring point is a weight of the corresponding modified radial basis function, is an error confidence, is a selected basis function, , and are normalized high-confidence neighboring points in temperature, humidity and power, respectively, is a high-confidence neighboring point normalization set, is a high-confidence neighboring point set is a medium-confidence neighboring point is a standard deviation of the high-confidence neighboring points, is a preset maximum error of a scene; and the function expression of the electric energy metering error correction value of the to-be-interpolated point is obtained by weighted summation of the weights of the modified radial basis functions and the electric energy metering error of the high-confidence neighboring point set . ； in, Points to be interpolated The power metering error correction value, For the set of high confidence neighbor points and the standardized set of high-confidence nearest neighbors Size, For the set of high confidence neighbor points Medium to high confidence neighboring points The corresponding weights of the modified radial basis function, For the set of high confidence neighbor points The error in electricity metering.

7. An electric vehicle charging pile electric energy metering error correction system comprising a microprocessor and a memory connected to each other, characterized in that, The microprocessor is programmed or configured to perform the electric vehicle charging pile electric energy measurement error correction method of any one of claims 1-6.

8. A computer-readable storage medium having stored therein a computer program or instructions, characterized in that, The computer program or instructions are programmed or configured to perform the electric vehicle charging pile electric energy measurement error correction method of any one of claims 1-6 by the processor.

9. A computer program product comprising computer programs or instructions, characterized in that, The computer program or instructions are programmed or configured to perform the electric vehicle charging pile electric energy measurement error correction method of any one of claims 1-6 by the processor.

Images