Electric energy metering method of intelligent electric energy metering box

Abstract

The invention discloses an electric energy metering method of an intelligent electric energy metering box, and relates to the technical field of electric power automation and intelligent metering, and the method comprises the following steps: collecting electric power load data generated in the operation process of power electronic equipment in real time, and carrying out the preprocessing of the collected data, so as to recognize and track the dynamic change characteristics of the equipment load. According to the invention, through real-time acquisition and analysis of equipment load data, dynamic calibration is carried out in combination with frequency spectrum transformation and a machine learning model, contribution factors of a high-order frequency band can be adaptively adjusted, and the influence of harmonic waves on electric energy is compensated. Meanwhile, the data sampling frequency is dynamically adjusted, it is ensured that the metering model responds to load fluctuation in real time, accuracy and timeliness are kept, error amplification caused by static fundamental wave calibration is avoided, and the accuracy and reliability of electric energy metering and the power quality monitoring effect are improved.

Description

Technical Field

[0001] The present invention relates to the technical fields of power automation and intelligent metering, and particularly to a power metering method for an intelligent electric energy metering box. Background Art

[0002] "Power metering of an intelligent electric energy metering box" refers to using an electric energy metering box integrated with intelligent modules (such as an electric energy acquisition unit, a data processing unit, a communication module, etc.) to collect and calculate electrical parameters such as voltage, current, and power of the connected load or electrical equipment in real time, and automatically measure and record the power consumption through built-in algorithms. At the same time, it has functions such as abnormal monitoring, remote communication, data uploading, and historical record tracing. Compared with traditional manual meter reading or basic meter metering, this method has higher accuracy, real-time performance, and automation level, and can realize intelligent supervision and energy efficiency optimization of users' electricity consumption behaviors, and is widely used in scenarios such as intelligent distribution networks, industrial parks, commercial buildings, and smart communities.

[0003] Meter calibration refers to the process of debugging and adjusting metering equipment (such as electric meters, current transformers, temperature sensors, etc.) to ensure that its measurement results are consistent with the standard or reference value. This process usually includes detecting, adjusting, comparing the measurement accuracy of the equipment, and confirming that the output data of the equipment meets the specified accuracy requirements. Calibration can be manual or automatic.

[0004] The role of meter calibration is to ensure that the measurement results of the equipment are accurate, reliable, and comply with national or industry metering standards. Through regular calibration, equipment deviations, faults, or aging problems can be detected in a timely manner, avoiding inaccurate power metering caused by measurement errors. For the power system, especially in the intelligent electric energy metering system, meter calibration can ensure the fairness and accuracy of electricity bill settlement, and improve the load dispatching and resource management efficiency of the power grid, avoiding financial losses, customer disputes, and systemic risks caused by inaccurate metering.

[0005] The existing technologies have the following deficiencies: In the process of power metering calibration of some existing intelligent electric energy metering boxes, the fundamental frequency (50Hz) of the power system is usually used as the only reference standard for calibration operations, and accuracy correction is only carried out for linear loads or ideal power signals. However, in actual applications, especially in scenarios such as industrial control systems, large data centers, and high-end commercial buildings, there are widely high-frequency harmonic source devices such as frequency converters, UPSs, and power rectifier devices. Such non-linear loads will generate rich high-order harmonic components during operation, resulting in obvious spectral expansion phenomena in the power grid voltage and current signals.

[0006] Due to the lack of the ability of the existing metering box calibration mechanism to identify and respond to the dynamic changes in different frequency characteristics, it is unable to perform adaptive parameter adjustment according to the harmonic components generated by the current load, resulting in sticking to the static fundamental wave calibration parameters in the load environment with significant harmonic changes. The resulting hysteresis of the metering model and the mismatch of the spectral response are likely to significantly amplify the power metering deviation, and the measurement error cannot be corrected in time. This problem not only affects the authenticity and accuracy of metering data during long-term operation, but may also have an adverse impact on electricity bill settlement, power quality analysis, and energy scheduling strategies.

[0007] The above information disclosed in the background art section is only used to enhance the understanding of the background of the present disclosure, and thus it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0008] The object of the present invention is to provide a power metering method for an intelligent power metering box. By collecting and analyzing the device load data in real time, and combining spectrum transformation and a machine learning model for dynamic calibration, it can adaptively adjust the contribution factor of the high-frequency band, compensate for the influence of harmonics on electric energy. At the same time, it dynamically adjusts the data sampling frequency to ensure that the metering model responds to load fluctuations in real time, maintains accuracy and timeliness, avoids the error amplification caused by static fundamental wave calibration, and improves the accuracy, reliability of power metering and the power quality monitoring effect, so as to solve the problems in the above background art.

[0009] To achieve the above object, the present invention provides the following technical solution: A power metering method for an intelligent power metering box, including the following steps: Collect the power load data generated by the power electronic device during operation in real time, and preprocess the collected data to identify and track the dynamic change characteristics of the device load; Based on the preprocessed load data, use the spectrum transformation method to decompose the signal in the frequency domain, construct a harmonic energy spectrogram reflecting the energy distribution, quantitatively calculate the frequency domain energy of different order harmonics, and obtain the energy proportion they account for in the overall power signal; By analyzing the harmonic energy spectrogram, extract the key indicators reflecting the high-order harmonic behavior of the power electronic device, comprehensively analyze the extracted key indicators, and quantitatively characterize the non-linear power response behavior of the power load; Input the analyzed key indicators into a pre-trained machine learning model, identify the non-linear behavior characteristics through the model, and determine the non-linear response level of the power electronic device in the current operating state; After identifying high - order harmonics generated during the operation of power electronic devices, based on the evaluation results, increase the contribution factor of the high - frequency band to compensate for its contribution to the total electrical energy. At the same time, dynamically adjust the data sampling frequency according to the load change, so that the metering model can track and adapt to the load fluctuation characteristics in real - time, ensuring the accuracy and real - time performance of metering.

[0010] Preferably, the specific steps for real - time collecting the power load data generated during the operation of power electronic devices are as follows: First, through high - precision sensors, electrically parameter signals during device operation are sensed in real - time; Second, convert the analog electrical signals into digital signals through a high - speed analog - to - digital converter and transmit them to the data acquisition module; Then, use the local processing unit to perform time synchronization, data caching, and preliminary verification on the collected data to ensure data continuity and validity; Finally, transmit the real - time data to the central control system through a wired or wireless communication module to provide basic data support for subsequent harmonic analysis, load identification, and power metering models.

[0011] Preferably, by analyzing the harmonic energy spectrum diagram, key indicators reflecting the high - order harmonic behavior of power electronic devices are extracted. The extracted indicators include the logarithmic difference between high - order harmonics and the fundamental wave and the energy density distribution of different frequency bands in the harmonic spectrum. Analyze the logarithmic difference between high - order harmonics and the fundamental wave and the energy density distribution of different frequency bands in the harmonic spectrum under the detection window to generate a harmonic logarithmic difference reference value and a harmonic spectrum density reference value respectively, and quantify the non - linear power response behavior of the power load through the harmonic logarithmic difference reference value and the harmonic spectrum density reference value.

[0012] Preferably, the specific steps for comprehensively analyzing the logarithmic difference between high - order harmonics and the fundamental wave under the detection window to generate a harmonic logarithmic difference reference value are as follows: Perform frequency - domain decomposition on the power signal through spectrum analysis to extract the amplitude values of the fundamental wave and each high - order harmonic. Let the amplitude of the fundamental wave be and the amplitude of the th high - order harmonic be , where , calculate the logarithmic difference between the fundamental wave and the high - order harmonic. The calculation expression is as follows: , where is the logarithmic difference between the th harmonic and the fundamental wave; Integrate all the logarithmic difference values of the harmonics to generate a harmonic logarithmic difference reference value for evaluating the non - linear load characteristics of power electronic devices. The calculation expression is as follows: , where ， is the reference value of the harmonic logarithm difference, is the weight coefficient of the th harmonic distribution,

[0013] Preferably, the specific steps for comprehensively analyzing the energy density distribution of different frequency bands in the harmonic spectrum to generate the harmonic spectrum density reference value are as follows: First, perform frequency-domain analysis on the load data generated by the power electronic device, and use the fast Fourier transform to convert the time-domain signal into a frequency-domain signal. In the frequency-domain signal, the amplitude of each frequency component reflects the energy distribution of each frequency band in the electrical energy signal. Through conversion, the energy density of each frequency band is obtained, and the calculation expression is as follows: , where is the energy density of the th frequency band, is the complex amplitude of the th frequency band obtained through spectrum analysis, is the frequency interval of the frequency band ; After obtaining the energy density of each frequency band, a harmonic spectrum density reference value is comprehensively generated. This reference value can comprehensively reflect the intensity of the high-order harmonics generated by the device. The generation formula of the harmonic spectrum density reference value is as follows: , where is the harmonic spectrum density reference value, is the weight function of the frequency band , is the total number of frequency bands participating in the calculation in the spectrum.

[0014] Preferably, input the analyzed harmonic logarithm difference reference value and harmonic spectrum density reference value into a pre-trained machine learning model, generate a high-order harmonic generation risk coefficient through the machine learning model, and identify the non-linear behavior characteristics through the high-order harmonic generation risk coefficient to determine the non-linear response level of the power electronic device under the current operating state.

[0015] Preferably, compare and analyze the high-order harmonic generation risk coefficient generated when identifying the non-linear behavior characteristics through the pre-trained machine learning model with the pre-set high-order harmonic generation risk coefficient reference threshold to determine the non-linear response level of the power electronic device under the current operating state. The judgment logic is as follows: If the high-order harmonic generation risk coefficient is greater than the pre-set high-order harmonic generation risk coefficient reference threshold, it is determined that high-order harmonics are generated during the operation of the power electronic device; if the high-order harmonic generation risk coefficient is less than or equal to the pre-set high-order harmonic generation risk coefficient reference threshold, it is determined that no high-order harmonics are generated during the operation of the power electronic device.

[0016] Preferably, after identifying high - order harmonics generated during the operation of the power electronic device, based on the evaluation results, the band contribution factor of the high - frequency band is increased to compensate for its contribution to the total electric energy. At the same time, the specific steps of dynamically adjusting the data sampling frequency according to the load change are as follows: Under the influence of high - order harmonics generated during the operation of the power electronic device, according to the evaluation results, the band contribution factor of the high - frequency band is increased to accurately compensate for its contribution to the total electric energy. The calculation formula for adjusting the band contribution factor is as follows: , where is the adjusted band contribution factor of the high - frequency band, is the band contribution factor of the original fundamental - wave band, is the high - order harmonic generation risk coefficient, is the adjustment coefficient; According to the dynamic change of the load, the data sampling frequency is dynamically adjusted. By adjusting the data sampling frequency, the electric energy metering system can track the load fluctuation in real - time to ensure high - precision electric energy metering when the load changes suddenly. The calculation expression is as follows: , where is the adjusted data sampling frequency, is the originally set basic sampling frequency, is the adjustment coefficient, is the change in load power, is the current load power; Track the fluctuation characteristics of the load in real - time, and dynamically adjust the electric energy metering model according to the high - order harmonic generation risk coefficient and load fluctuation to ensure the metering accuracy and real - time performance under the conditions of load fluctuation and high - order harmonic generation. The calculation expression is as follows: , where is the adjusted electric energy metering result, is the original electric energy metering result, is the high - order harmonic adjustment factor, is the load fluctuation adjustment factor, is the change in sampling frequency, and the calculation formula is as follows: .

[0017] In the above technical solution, the technical effects and advantages provided by the present invention are: By collecting and analyzing the load data during the operation of the device in real time, the present invention accurately identifies and tracks the dynamic change characteristics of the load, combines spectrum transformation and machine learning models for dynamic calibration, and can adaptively adjust the contribution factor of the high-frequency band, thereby compensating for the impact of high-order harmonics on the total electric energy. At the same time, by dynamically adjusting the data sampling frequency, the metering model can respond to load fluctuations in real time, ensuring the accuracy and timeliness of measurement under high-order harmonic load environments, thus avoiding the error amplification problem caused by traditional static fundamental wave calibration methods and ensuring the accuracy, reliability of electric energy metering and the monitoring effect of power quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required in the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention, and those of ordinary skill in the art can also obtain other drawings based on these drawings.

[0019] Figure 1 It is a flowchart of the electric energy metering method of the intelligent electric energy metering box of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] Now, the exemplary embodiments will be described more fully with reference to the accompanying drawings. However, the exemplary embodiments can be implemented in various forms and should not be construed as limited to the examples set forth herein; rather, these exemplary embodiments are provided so that this disclosure will be more complete and comprehensive, and will fully convey the concept of the exemplary embodiments to those skilled in the art.

[0021] The present invention provides an electric energy metering method for an intelligent electric energy metering box as Figure 1 shown, including the following steps: Collect in real time the power load data generated during the operation of power electronic devices (such as frequency converters, UPSs, etc.), and preprocess the collected data to identify and track the dynamic change characteristics of the device load; Continuously monitor the current and voltage signals through high-precision sensors installed in the electric energy metering box. These data include not only the current and voltage of the fundamental wave (50 Hz), but also the changes in the high-order harmonic components. The collected load data will be used as the basis for subsequent processing to ensure that all load fluctuations and harmonic interferences are comprehensively considered.

[0022] The raw data collected in real time often contains noise (such as electromagnetic interference, transmission delay, etc.), so filtering and denoising processes are required to ensure the accurate extraction of harmonic components. For example, common preprocessing methods include low-pass, high-pass filters, detrending, etc. In addition, standardizing the data is also a necessary step to eliminate errors caused by range differences or inconsistent units, ensuring the unity and accuracy of subsequent calculations.

[0023] By obtaining in real time the electrical energy data generated during the operation of power electronic devices (such as frequency converters, UPS, etc.), including various electrical parameters such as voltage, current, power, etc., the purpose is to monitor the working status and load characteristics of the devices. The collected data will go through a preprocessing process, which includes steps such as denoising, filtering, and standardization, so as to ensure the accuracy and consistency of the data. Through these processes, the system can identify the dynamic change characteristics of the device load, such as load fluctuations, nonlinear characteristics, and harmonic generation. The core of this process is to be able to reflect the changes in the device operation status in real time, helping to further analyze and optimize the operation efficiency and power consumption of the device.

[0024] The specific steps for real-time collection of the electrical load data generated during the operation of power electronic devices are as follows: First, through high-precision sensors (such as voltage and current transformers), the electrical parameter signals during device operation are sensed in real time; Second, the analog electrical signals are converted into digital signals through a high-speed analog-to-digital converter (ADC) and transmitted to the data acquisition module; Then, a local processing unit (such as an embedded chip or an edge computing module) is used to perform time synchronization, data caching, and preliminary verification on the collected data to ensure data continuity and validity; Finally, the real-time data is transmitted to the central control system through a wired or wireless communication module (such as RS485, CAN, Ethernet, Wi-Fi, etc.), providing basic data support for subsequent harmonic analysis, load identification, and power metering models.

[0025] Based on the preprocessed load data, the spectral transformation method is used to decompose the signal in the frequency domain, construct a harmonic energy spectrum diagram reflecting the energy distribution, quantitatively calculate the frequency domain energy of different order harmonics, and obtain the proportion of the energy they account for in the overall electrical signal; Based on the preprocessed load data (such as signals like voltage, current, etc.), the frequency-domain analysis of the signal is carried out using spectral transformation methods (such as discrete Fourier transform (DFT) or fast Fourier transform (FFT)), converting the signal in the time domain into a signal in the frequency domain. Through frequency-domain decomposition, the original signal can be decomposed into the superposition of different frequency components, identifying the fundamental frequency (50Hz) and various high-order harmonic components (such as 150Hz, 250Hz, etc.). Then, a harmonic energy spectrum diagram is constructed, which shows the energy distribution of different frequency components. Usually, the frequency is taken as the abscissa and the signal amplitude or power as the ordinate, reflecting the contribution of each frequency band to the total power consumption. In this way, the harmonic components in the signal and their energy intensities can be visually displayed, helping to analyze the harmonics generated by the equipment and their impact on power consumption.

[0026] The role of the harmonic energy spectrum diagram is to visually display the energy distribution of each frequency component (including the fundamental wave and high-order harmonics) in the total power consumption of the power system. Through the frequency-domain analysis of the load signal, the harmonic energy spectrum diagram can help identify the harmonic components of different frequencies generated during the operation of the equipment or system, revealing the contribution of high-order harmonics to the total power consumption. It provides the energy intensity information of each frequency band, enabling engineers to analyze the degree of harmonic pollution caused by power electronic devices (such as frequency converters, UPS, etc.), and providing decision-making support for power quality assessment, the design of harmonic compensation devices, and the optimization of the operation of the power system. Through this analysis, the stability and efficiency of the power system can be effectively evaluated, avoiding measurement errors or equipment failures caused by harmonic mismatch.

[0027] By analyzing the harmonic energy spectrum diagram, key indicators reflecting the high-order harmonic behavior of power electronic devices are extracted, and a comprehensive analysis of the extracted key indicators is carried out to quantify the non-linear power response behavior of the power load; By analyzing the harmonic energy spectrum diagram, key indicators reflecting the high-order harmonic behavior of power electronic devices are extracted. The extracted indicators include the logarithmic difference between the high-order harmonics and the fundamental wave and the energy density distribution of different frequency bands in the harmonic spectrum. The logarithmic difference between the high-order harmonics and the fundamental wave and the energy density distribution of different frequency bands in the harmonic spectrum are analyzed under the detection window to generate the harmonic logarithmic difference reference value and the harmonic spectrum density reference value respectively. The non-linear power response behavior of the power load is quantified through the harmonic logarithmic difference reference value and the harmonic spectrum density reference value; When the logarithmic difference between the high - order harmonics and the fundamental wave becomes larger, it indicates that the power electronic equipment generates strong high - order harmonics during operation. This is because power electronic equipment (such as frequency converters, UPS, etc.) usually introduces non - linear loads during operation, resulting in distortion of current and voltage waveforms. The fundamental wave (50Hz) usually represents the main signal component of the power system, while the high - order harmonics are integer - multiple frequency components of the fundamental wave. When strong high - order harmonics are generated during equipment operation, their amplitude ratio to the fundamental wave increases, thus increasing their logarithmic difference. This means that the energy distribution of the harmonics has changed significantly, and the contribution of high - order harmonics gradually occupies a larger proportion of the power consumption, and may even cause instability and efficiency loss of the power system. Specifically, the increase in the logarithmic difference reflects the increase in the energy proportion of high - order harmonics in the spectrum, indicating that the non - linear characteristics of the load are intensified, and the power electronic equipment frequently generates or strengthens high - order harmonics during operation. This change not only affects the accuracy of power metering, but also may lead to equipment overheating, power quality problems, and other non - linear phenomena in the system. Therefore, the increase in the logarithmic difference between high - order harmonics and the fundamental wave is actually a strong indication of the generation of high - order harmonics by power electronic equipment.

[0028] The specific steps for comprehensively analyzing the logarithmic difference between high - order harmonics and the fundamental wave under the detection window to generate a harmonic logarithmic difference reference value are as follows: Through frequency - domain decomposition of the power signal by spectrum analysis (such as FFT), extract the amplitude values of the fundamental wave (50Hz) and each high - order harmonic (such as the 3rd harmonic, 5th harmonic, etc.). Let the amplitude of the fundamental wave be , and the amplitude of the th high - order harmonic be . Among them, calculate the logarithmic difference between the fundamental wave and the high - order harmonic. The calculation expression is as follows: , where is the logarithmic difference between the th harmonic and the fundamental wave; In this step, by comparing the ratio of the amplitude of the high - order harmonic to the amplitude of the fundamental wave and taking its logarithm, the generation intensity of the high - order harmonic relative to the fundamental wave is quantified. The role of the logarithmic operation is to compress the quantification range of the amplitude difference, so that larger differences can be more clearly displayed. Here, represents the logarithmic difference between the th harmonic and the fundamental wave. The larger the logarithmic difference value, the more significant the contribution of the high - order harmonic compared to the fundamental wave, indicating that the intensity of the high - order harmonics generated by the power electronic equipment during operation is greater.

[0029] Sum up the logarithmic difference values of all harmonics Perform synthesis to generate the reference value of the harmonic logarithm difference for evaluating the non-linear load characteristics of power electronic devices. To this end, the logarithmic difference values of different harmonic components are weighted and averaged to obtain the final reference value. The calculation expression is as follows: , where ， is the reference value of the harmonic logarithm difference, is the weight coefficient assigned to the th harmonic, which is used to reflect the importance of this harmonic component in the total power consumption, is the maximum order of the harmonics, that is, the order of the highest harmonic considered when calculating the harmonic logarithm difference index.

[0030] Combine the logarithmic difference values between each high-order harmonic and the fundamental wave to form a reference value of the harmonic logarithm difference, so as to comprehensively evaluate the high-order harmonic generation intensity of power electronic devices. By weighted accumulation of the logarithmic differences of each harmonic, the index value can accurately reflect the non-linearity degree of the device load, helping to identify the influence of high-order harmonics and their contributions in the total power consumption.

[0031] The larger the reference value of the harmonic logarithm difference generated by comprehensively analyzing the logarithmic difference between the high-order harmonic and the fundamental wave under the detection window, the stronger the high-order harmonics generated by the power electronic device during operation. The reference value of the harmonic logarithm difference quantifies the non-linear characteristics of the device load by comparing the logarithmic difference between the high-order harmonic and the fundamental wave. When power electronic devices (such as frequency converters, UPSs, etc.) are operating, when the amplitude ratio of the generated high-order harmonic components to the fundamental wave increases, the logarithmic difference of the harmonics will increase significantly. This increased logarithmic difference reflects the increased contribution of the high-order harmonics in the total power consumption, meaning that the device generates stronger high-order harmonics, resulting in an increase in the performance value of the reference value of the harmonic logarithm difference. On the contrary, when this reference value is small or close to zero, it indicates that the device load is mainly composed of the fundamental wave, the generation of high-order harmonics is relatively weak, and the non-linear characteristics during the device operation are low. Therefore, a small or near-zero value indicates that no significant high-order harmonics are generated during the operation of the power electronic device.

[0032] An increase in the energy density of different frequency bands in the harmonic spectrum usually indicates that high - order harmonics are generated during the operation of power electronic devices. This is because nonlinear power electronic devices (such as frequency converters, UPSs, power rectifiers, etc.) introduce a large number of non - sine wave components during switching, rectification, or modulation, thus generating harmonic frequencies other than the fundamental wave in the spectrum. When the energy density of these high - order harmonic frequency bands (such as 150Hz, 250Hz, 350Hz, etc.) increases significantly, it means that the amplitudes of these frequency components become larger, that is, the "proportion" of the harmonic signal in the spectrum increases, reflecting an enhanced nonlinearity of the device under the current operating state. This phenomenon indicates that the internal control strategy or workload of the power electronic device has changed, resulting in an increase in high - frequency electrical disturbances, which is a direct manifestation of the generation of high - order harmonics. The increase in harmonic energy density not only indicates the appearance of harmonics but also reveals an increase in their electrical energy contribution, becoming an important basis for judging high - order harmonic behavior. Therefore, by monitoring the change in the energy density of different frequency bands in the spectrum, it is possible to accurately identify whether the power electronic device is in a state of enhanced harmonic generation, which is a key reference index in power quality analysis and the formulation of harmonic suppression strategies.

[0033] The specific steps for comprehensively analyzing the energy density distribution of different frequency bands in the harmonic spectrum to generate a reference value of harmonic spectrum density under the detection window are as follows: First, perform frequency - domain analysis on the load data generated by the power electronic device. Use the fast Fourier transform (FFT) to convert the time - domain signal into a frequency - domain signal. In the frequency - domain signal, the amplitude of each frequency component reflects the energy distribution of each frequency band in the electrical energy signal. Through conversion, the energy density of each frequency band is obtained, and the calculation formula is as follows: , where is the energy density of the th frequency band, is the complex amplitude (i.e., the amplitude of the corresponding frequency band) of the th frequency band obtained through spectrum analysis, reflecting the signal strength of this frequency band, is the frequency interval of the frequency band ; The function of this step is to calculate the energy density of each frequency band, which reflects the contribution of different frequency components to the total signal energy. By calculating the energy density of each frequency band, the "strength" or "activity" of each frequency point in the spectrum can be obtained, thereby revealing the harmonic signals generated by the device in this frequency band.

[0034] After obtaining the energy density of each frequency band, a reference value of harmonic spectrum density is comprehensively generated. This reference value can comprehensively reflect the intensity of high - order harmonics generated by the device. To emphasize the contribution of high - order harmonics, higher weights are usually assigned to the high - frequency part of the spectrum (such as the 2nd, 3rd, 5th harmonics exceeding the fundamental wave). The formula for generating the reference value of harmonic spectrum density is as follows: , where is the reference value of the harmonic spectrum density, is the frequency band of the weight function, which is used to adjust the contribution of each frequency band in the final calculation, is the total number of frequency bands participating in the calculation in the spectrum, indicating the frequency range we analyze.

[0035] In this step, a comprehensive reference value of the harmonic spectrum density is obtained by weighted summation of the energy density of each frequency band. The weight function can be designed and adjusted according to the characteristics of the device or the requirements for harmonic contribution. In this way, the generation situation of high-frequency band harmonics is highlighted, which helps to judge whether high-order harmonics are generated and their intensity during the operation of power electronic devices, thus providing a basis for power quality analysis.

[0036] The larger the reference value of the harmonic spectrum density generated after comprehensively analyzing the energy density distribution of different frequency bands in the harmonic spectrum under the detection window, it indicates that there is significant energy density accumulation in multiple high-order harmonic frequency bands (such as 150Hz, 250Hz, 350Hz, etc.), which means that the amplitude or power proportion of harmonic components in these frequency bands is relatively high, reflecting the enhanced nonlinear characteristics during the operation of power electronic devices and the large generation of high-order harmonics. On the contrary, when the reference value is small, it indicates that the energy density in the high-frequency band is low, and the spectral energy is mainly concentrated in the fundamental wave or low-order harmonic part, indicating that the device is operating relatively smoothly and no obvious high-order harmonics are generated.

[0037] Input the analyzed key indicators into a pre-trained machine learning model, and identify the nonlinear behavior characteristics through the model to determine the nonlinear response level of the power electronic device under the current operating state; Input the analyzed harmonic logarithm difference reference value and the reference value of the harmonic spectrum density into a pre-trained machine learning model, generate a high-order harmonic generation risk coefficient through the machine learning model, and identify the nonlinear behavior characteristics through the high-order harmonic generation risk coefficient to determine the nonlinear response level of the power electronic device under the current operating state.

[0038] Judging the nonlinear response level of the power electronic device under the current operating state refers to evaluating whether the device shows nonlinear load characteristics during operation and the impact degree of this nonlinear load on the power system. Specifically, when power electronic devices (such as frequency converters, UPSs, power rectification devices, etc.) are operating, the current or voltage waveforms may be distorted due to the working mode of the internal circuit (such as switching actions, signal rectification, inversion, etc.), generating high-order harmonics. These high-order harmonic components reflect the degree of nonlinear response of the device.

[0039] By analyzing the electrical energy signals of the device, especially by identifying the generation characteristics of high - order harmonics, the non - linear response level of the device can be quantified. If the high - order harmonics generated by the device are significant and do not conform to the ideal linear waveform, it indicates that the non - linear response of the device is strong, which may lead to power quality problems such as harmonic pollution and power factor decline. Therefore, judging the non - linear response level is not only to detect the generation of high - order harmonics, but more importantly, to evaluate the potential impact of these harmonics on the power grid, so as to take necessary compensation measures or optimize power dispatching.

[0040] A pre - trained machine - learning model refers to a machine - learning algorithm model that has completed the training process using a large amount of historical sample data before being formally applied to the electric energy metering system. It learns the electrical characteristics of the device under different working conditions (including harmonic indexes, load responses, waveform distortion, etc.) and establishes a mapping relationship between input features and target results. In this scenario, the model usually mainly uses supervised learning algorithms. During the training process, a large number of labeled sample data are input, and the actual non - linear degree or harmonic risk level is used as the label. By continuously iteratively optimizing the parameters, the model learns how to extract key factors from multiple complex features to predict or evaluate the high - order harmonic generation risk in an unknown state. The pre - training process may include multiple stages such as feature selection, normalization processing, model structure design (such as the number of neural network layers, the type of support vector machine kernel function, etc.), and cross - validation, ensuring that the model has good generalization ability and judgment ability when facing new data.

[0041] In the application stage, this model is deployed in the embedded processing platform or edge server of the intelligent electric energy metering box to judge the harmonic feature data (such as the reference value of harmonic logarithm difference and the reference value of harmonic spectral density) obtained by real - time acquisition and analysis. When these data are input into the model, the model calculates a risk coefficient representing the high - order harmonic generation risk, that is, the possibility or intensity level of high - order harmonic generation, according to the feature patterns it has learned. This risk coefficient is further used to identify whether the current device is in a non - linear operating state and to judge whether its non - linear response level is low, medium or high. Since the model has been trained in a large number of device scenarios and can comprehensively consider the multi - dimensional feature coupling relationship under complex working conditions, it is more intelligent and flexible than the traditional fixed - threshold judgment method and is especially suitable for complex and highly volatile industrial electricity scenarios. Introducing this model into the intelligent metering system not only improves the ability to identify high - order harmonic generation behavior, but also significantly enhances the dynamic perception and adaptive adjustment ability of the entire system to non - linear load behavior.

[0042] The machine - learning model is not limited here, and it can realize the comprehensive analysis of the reference value of harmonic logarithm difference and the reference value of harmonic spectral density to generate a high - order harmonic generation risk coefficient Any machine learning model can be used. To implement the technical solution of the present invention, the present invention provides a specific implementation method: High-order harmonic generation risk coefficient The generation formula is as follows: , where in the formula, and are respectively the reference values of the harmonic logarithm difference and the reference value of the harmonic spectrum density preset proportionality coefficients, and and are both greater than 0.

[0043] The preset proportionality coefficients (i.e., and in the formula) refer to the weighting factors used to perform weighted calculations on the harmonic logarithm difference reference value and the harmonic spectrum density reference value of two different reference indicators when generating the high-order harmonic generation risk coefficient . They respectively represent the relative importance or influence degree of the two reference values in the overall risk coefficient calculation, and are set artificially in a preset manner or given according to experience, historical data or scenario weights during the system initialization stage.

[0044] In other words, and are the adjustment parameters during the normalization fusion of the input features, used to control the proportion of the harmonic logarithm difference reference value and the harmonic spectrum density reference value when synthesizing the final risk coefficient . For example, if the harmonic spectrum density can more accurately reflect the harmonic behavior in a certain type of power system, can be set to enhance the influence of . Since both of these coefficients are positive numbers greater than 0, they can also avoid the abnormal situation of the denominator being zero in mathematical calculations, thus ensuring the stability and adjustability of the entire model calculation.

[0045] It can be seen from the high-order harmonic generation risk coefficient that the larger the reference value of the harmonic logarithm difference generated by comprehensively analyzing the logarithmic difference between the high-order harmonic and the fundamental wave under the detection window, and the larger the reference value of the harmonic spectrum density generated by comprehensively analyzing the energy density distribution of different frequency bands in the harmonic spectrum under the detection window, the greater the high-order harmonic generation risk coefficient generated when identifying the non-linear behavior characteristics through the pre-trained machine learning model, indicating that the probability of generating high-order harmonics during the operation of the power electronic device is greater. On the contrary, it indicates that the probability of generating high-order harmonics during the operation of the power electronic device is smaller.

[0046] Compare and analyze the high - order harmonic generation risk coefficient generated when identifying non - linear behavior characteristics through a pre - trained machine learning model with the pre - set reference threshold of the high - order harmonic generation risk coefficient to determine the non - linear response level of the power electronic device under the current operating state. The judgment logic is as follows: If the high - order harmonic generation risk coefficient is greater than the pre - set reference threshold of the high - order harmonic generation risk coefficient, it is determined that high - order harmonics are generated during the operation of the power electronic device; if the high - order harmonic generation risk coefficient is less than or equal to the pre - set reference threshold of the high - order harmonic generation risk coefficient, it is determined that no high - order harmonics are generated during the operation of the power electronic device.

[0047] After identifying that the power electronic device generates high - order harmonics during operation, based on the evaluation results, increase the frequency - band contribution factor of the high - order frequency band to compensate its contribution to the total electrical energy. At the same time, dynamically adjust the data sampling frequency according to the load change, so that the metering model can track and adapt to the load fluctuation characteristics in real time, ensuring the accuracy and real - time performance of metering; The function of this step is to realize the dynamic identification and adaptive calibration compensation of the high - order harmonics generated during the operation of the power electronic device, so as to improve the metering accuracy and real - time response ability of the intelligent power metering system in a complex load environment. Specifically, when power electronic devices (such as frequency converters, UPSs, rectifiers, etc.) generate high - order harmonics during operation, these harmonics will significantly affect the spectral structure of the voltage and current waveforms, resulting in the inability of traditional static power metering methods based on the fundamental frequency (50Hz) reference to accurately reflect the true power consumption of each frequency band. Therefore, by increasing the frequency - band contribution factor of the high - order frequency band, the calculation weights of each frequency component for the total electrical energy can be redistributed in the metering model, ensuring that the power consumption of high - frequency harmonic components will not be ignored or underestimated, thereby compensating for the metering deviation caused by harmonics.

[0048] In addition, this step also includes dynamically adjusting the data sampling frequency according to the load fluctuation of the device, that is, automatically increasing or decreasing the data sampling rate according to the harmonic fluctuation characteristics, so as to improve the response sensitivity of the metering system to high - frequency changes. Through this dynamic matching of the sampling frequency, it can be ensured that when the harmonic frequency changes rapidly or the non - linear load mutates, the system can still obtain signal features with sufficient resolution, preventing high - frequency energy "aliasing" or being missed. Finally, through the "high - frequency weighted compensation + dynamic sampling response" collaborative mechanism, the power metering model has stronger adaptability and real - time tracking ability, can stably output accurate metering results in high - harmonic scenarios, and significantly improve the reliability and accuracy of the system in practical applications.

[0049] After identifying that the power electronic device generates high - order harmonics during operation, based on the evaluation results, the specific steps of increasing the frequency - band contribution factor of the high - order frequency band to compensate its contribution to the total electrical energy and dynamically adjusting the data sampling frequency according to the load change are as follows: Under the influence of high - order harmonics generated during the operation of power electronic devices, the band contribution factor in the high - order frequency band is increased according to the evaluation results to accurately compensate for its contribution to the total electrical energy. To adjust the impact of the high - order frequency band on electrical energy metering, the size of the band contribution factor is adjusted according to the high - order harmonic generation risk coefficient, so as to ensure that the electrical energy contribution of these frequency bands is accurately reflected. The calculation formula for adjusting the band contribution factor is as follows: , where is the adjusted band contribution factor of the high - order frequency band, is the band contribution factor of the original fundamental frequency band (i.e., the contribution factor of the 50Hz fundamental wave), is the high - order harmonic generation risk coefficient, is the adjustment coefficient, which represents the adjustment amplitude of the band contribution factor in the high - order frequency band, used to adjust the size of the band contribution factor in the high - order frequency band, and is a coefficient for flexibly controlling the weight of the high - order frequency band; The function of this step is to make up for the impact of high - order harmonics on the total electrical energy by increasing the weight of the high - order frequency band, so as to ensure the accuracy of the electrical energy metering result and avoid measurement errors caused by not considering the harmonic components.

[0050] According to the dynamic changes of the load, especially the fluctuations of high - order harmonics, the data sampling frequency is dynamically adjusted. By adjusting the data sampling frequency, the electrical energy metering system can track the load fluctuations in real - time, ensuring high - precision electrical energy metering when the load changes suddenly. The calculation expression is as follows: , where is the adjusted data sampling frequency, is the original set basic sampling frequency, is the adjustment coefficient, controlling the sensitivity of the sampling frequency change, is the change in load power, is the current load power; The function of this step is to dynamically adjust the data sampling frequency according to the changes of the load, especially the changes in load power and the intensity of high - order harmonics. This can ensure that the system can respond to the load fluctuations in real - time, thus obtaining more accurate electrical energy metering data.

[0051] Track the fluctuation characteristics of the load in real - time, and dynamically adjust the electrical energy metering model according to the high - order harmonic generation risk coefficient and load fluctuations, ensuring the measurement accuracy and real - time performance under the conditions of load fluctuations and high - order harmonic generation. The calculation expression is as follows: , where is the adjusted electrical energy metering result, representing the electrical energy metering value after dynamic adjustment considering high - order harmonics and load fluctuations, is the original electrical energy metering result, is the high - order harmonic adjustment factor, which controls the influence degree of high - order harmonics on the electricity metering result. is the load fluctuation adjustment factor, which controls the influence degree of load fluctuation on the electricity metering result. is the change amount of sampling frequency, and the calculation formula is as follows: .

[0052] The function of this step is to combine the high - order harmonic generation risk factor and dynamic sampling frequency adjustment to ensure that the electricity metering system can adapt to load fluctuations and high - order harmonic generation in real time, so as to maintain the accuracy and real - time performance of metering. Especially in the case of high - harmonic loads and frequent load fluctuations, the accuracy of the electricity metering result can be ensured.

[0053] The present invention can accurately identify and track the dynamic change characteristics of the load by collecting and analyzing the load data during the operation of the device in real time, and perform dynamic calibration by combining spectrum transformation and machine - learning models. It can adaptively adjust the contribution factor of the high - order frequency band to compensate for the influence of high - order harmonics on the total electric energy. At the same time, by dynamically adjusting the data sampling frequency, the metering model can respond to load fluctuations in real time, ensuring the accuracy and timeliness of measurement in a high - order harmonic load environment, thus avoiding the error amplification problem brought by the traditional static fundamental wave calibration method and ensuring the accuracy, reliability of electricity metering and the monitoring effect of power quality.

[0054] The above formulas are all calculated by taking the numerical value without dimension. The formulas are obtained by collecting a large amount of data for software simulation to get a formula closest to the real situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0055] Only some exemplary embodiments of the present invention are described by way of illustration above. Undoubtedly, for those of ordinary skill in the art, the described embodiments can be modified in various different ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the protection scope of the claims of the present invention.

[0056] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprising", "including" or any other variant thereof are intended to cover non-exclusive inclusion, such that a process, method, article or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation, an element defined by the phrase "comprising an..." does not exclude the presence of additional identical elements in the process, method, article or device comprising the element.

[0057] It should be understood that in various embodiments of the present application, the magnitudes of the serial numbers of the above processes do not imply the order of execution, and the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0058] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. A professional technician can use different methods for each specific application to implement the described functions, but such implementation should not be considered to exceed the scope of the present application.

[0059] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated herein.

[0060] The unit described as a separate component may or may not be physically separated, and the component shown as a unit may or may not be a physical unit, that is, it may be located in one place, or may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0061] In addition, in each embodiment of the present application, the functional units can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit.

[0062] As described above, it is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims described above.

[0063] Only some exemplary embodiments of the present invention have been described above by way of illustration. Undoubtedly, for those of ordinary skill in the art, the described embodiments can be modified in various different ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and description are illustrative in nature and should not be construed as limiting the protection scope of the claims of the present invention.

Claims

1. The electric energy metering method of the intelligent electric energy metering box is characterized in that: The following steps are involved: Collect the power load data generated by power electronic equipment during operation in real time, and pre-process the collected data to identify and track the dynamic change characteristics of equipment load; Based on the preprocessed load data, the signal is decomposed in the frequency domain using the spectrum transformation method, a harmonic energy spectrum reflecting the energy distribution is constructed, and the frequency domain energy of harmonics of different orders is quantified to obtain their energy proportion in the overall power signal. By analyzing the harmonic energy spectrum, key indicators reflecting the high-order harmonic behavior of power electronic equipment are extracted, and the extracted key indicators are comprehensively analyzed to quantify the nonlinear power response behavior of the power load; The analyzed key indicators are input into the pre-trained machine learning model, and the nonlinear behavior characteristics are identified through the model to determine the nonlinear response level of the power electronic equipment under the current operating state; When it is identified that power electronic equipment generates high-order harmonics during operation, the frequency band contribution factor of the high-order frequency band is increased based on the evaluation results to compensate for its contribution to the total electric energy. At the same time, the data sampling frequency is dynamically adjusted according to load changes, so that the metering model can track and adapt to load fluctuation characteristics in real time, ensuring the accuracy and real-time performance of the metering.

2. The electric energy metering method of the intelligent electric energy metering box according to claim 1 is characterized in that: The specific steps for real-time collection of power load data generated by power electronic equipment during operation are as follows: First, high-precision sensors are used to sense the electrical parameter signals of the equipment in real time during operation; Secondly, the analog electrical signal is converted into a digital signal through a high-speed analog-to-digital converter and transmitted to the data acquisition module; Next, the local processing unit is used to perform time synchronization, data caching and preliminary verification of the collected data to ensure data continuity and validity; Finally, the real-time data is transmitted to the central control system through wired or wireless communication modules, providing basic data support for subsequent harmonic analysis, load identification and power metering models.

3. The electric energy metering method of the intelligent electric energy metering box according to claim 1 is characterized in that: By analyzing the harmonic energy spectrum, key indicators reflecting the high-order harmonic behavior of power electronic equipment are extracted. The extracted indicators include the logarithmic difference between high-order harmonics and fundamental waves and the energy density distribution in different frequency bands in the harmonic spectrum. The logarithmic difference between high-order harmonics and fundamental waves and the energy density distribution in different frequency bands in the harmonic spectrum are analyzed under the detection window to generate harmonic logarithmic difference reference values ​​and harmonic spectrum density reference values ​​respectively. The nonlinear electric energy response behavior of the power load is quantified by the harmonic logarithmic difference reference values ​​and harmonic spectrum density reference values.

4. The electric energy metering method of the intelligent electric energy metering box according to claim 3 is characterized in that: The specific steps of comprehensively analyzing the logarithmic difference between the higher harmonics and the fundamental wave in the detection window to generate the reference value of the harmonic logarithmic difference are as follows: The power signal is decomposed in the frequency domain through spectrum analysis to extract the amplitude values ​​of the fundamental wave and each higher harmonic. The amplitude of the fundamental wave is set to , and the higher harmonics The subharmonic amplitude is ,in , calculate the logarithmic difference between the fundamental wave and the higher harmonics, the calculation expression is as follows: , where It is The logarithmic difference between the subharmonics and the fundamental; The logarithmic difference of all harmonics The harmonic logarithmic difference reference value is synthesized to evaluate the nonlinear load characteristics of power electronic equipment. The calculation expression is as follows: , where ， is the harmonic logarithmic difference reference value, It is The weight coefficient of subharmonic distribution, is the maximum order of the harmonics.

5. The electric energy metering method of the intelligent electric energy metering box according to claim 3 is characterized in that: The specific steps of comprehensively analyzing the energy density distribution of different frequency bands in the harmonic spectrum under the detection window to generate the harmonic spectrum density reference value are as follows: First, the load data generated by the power electronic equipment is analyzed in the frequency domain. The time domain signal is converted into the frequency domain signal using the fast Fourier transform. In the frequency domain signal, the amplitude of each frequency component reflects the energy distribution of each frequency band in the electric energy signal. Through the conversion, the energy density of each frequency band is obtained. The calculation expression is as follows: , where It is The energy density of each frequency band, It is obtained through spectrum analysis. The complex amplitude of the frequency band, It is the frequency band The frequency interval of After obtaining the energy density of each frequency band, the harmonic spectrum density reference value is comprehensively generated. This reference value can comprehensively reflect the intensity of high-order harmonics generated by the equipment. The formula for generating the harmonic spectrum density reference value is as follows: , where is the harmonic spectral density reference value, It is the frequency band The weight function of is the total number of frequency bands in the spectrum that participate in the calculation.

6. The electric energy metering method of the intelligent electric energy metering box according to claim 3 is characterized in that: The analyzed harmonic logarithmic difference reference value and harmonic spectrum density reference value are input into a pre-trained machine learning model, and a high-order harmonic generation risk coefficient is generated by the machine learning model. The nonlinear behavior characteristics are identified by the high-order harmonic generation risk coefficient, and the nonlinear response level of the power electronic equipment under the current operating state is determined.

7. The electric energy metering method of the intelligent electric energy metering box according to claim 6, characterized in that: The high-order wave generation risk coefficient generated when the pre-trained machine learning model identifies the nonlinear behavior characteristics is compared and analyzed with the pre-set high-order wave generation risk coefficient reference threshold to determine the nonlinear response level of the power electronic equipment under the current operating state. The judgment logic is as follows; If the high-order wave generation risk coefficient is greater than the preset high-order wave generation risk coefficient reference threshold value, it is judged that high-order wave is generated during the operation of the power electronic equipment; if the high-order wave generation risk coefficient is less than or equal to the preset high-order wave generation risk coefficient reference threshold value, it is judged that no high-order wave is generated during the operation of the power electronic equipment.

8. The electric energy metering method of the intelligent electric energy metering box according to claim 7, characterized in that: When it is identified that power electronic equipment generates high-order harmonics during operation, based on the evaluation results, the frequency band contribution factor of the high-order frequency band is increased to compensate for its contribution to the total electric energy. At the same time, the specific steps of dynamically adjusting the data sampling frequency according to load changes are as follows: Under the influence of high-order harmonics generated during the operation of power electronic equipment, the frequency band contribution factor of the high-order frequency band is increased according to the evaluation results to accurately compensate for its contribution to the total electric energy. The frequency band contribution factor adjustment calculation formula is as follows: , where is the band contribution factor of the adjusted higher-order frequency band, is the band contribution factor of the original fundamental frequency band, is the risk factor of high-order harmonic generation, is the adjustment coefficient; According to the dynamic changes of the load, the data sampling frequency is dynamically adjusted. By adjusting the data sampling frequency, the electric energy metering system can track the load fluctuation in real time to ensure high-precision electric energy metering when the load changes suddenly. The calculation expression is as follows: , where is the adjusted data sampling frequency, is the original basic sampling frequency, is the adjustment coefficient, is the load power change, is the current load power; Track the fluctuation characteristics of the load in real time, and dynamically adjust the electric energy metering model according to the risk factor of high-order harmonic generation and load fluctuation to ensure that the metering accuracy and real-time performance are maintained under the condition of load fluctuation and high-order harmonic generation. The calculation expression is as follows: , where is the adjusted energy measurement result, is the original electric energy measurement result, is the higher harmonic adjustment factor, is the load fluctuation adjustment factor, is the change in sampling frequency, and the calculation formula is as follows: .

Images