Method and device for electric energy metering of grid-connected new energy

Abstract

The application discloses a method and device for electric energy metering of grid-connected new energy, wherein wide-frequency harmonic characteristic information and inter-harmonic characteristic information are obtained according to pretreated grid voltage and grid current; after the wide-frequency harmonic characteristic information and the inter-harmonic characteristic information are input into an influence analysis topology network, first error compensation parameters are calculated by using error calculation of electric energy metering output by the influence analysis topology network; based on a non-sinusoidal signal simulation model, non-sinusoidal signals are obtained by using the pretreated grid voltage and the grid current combined with a ladder wave signal; second error compensation parameters are calculated by performing error compensation on the non-sinusoidal signals; time-varying error data of an electric energy meter are obtained by using historical operation data of the electric energy meter; target error compensation parameters are calculated based on the first error compensation parameters, the second error compensation parameters and the time-varying error data of the electric energy meter; electric energy metering is performed by the electric energy meter based on the target error compensation parameters, electric energy metering results are obtained, and the accuracy of electric energy metering of grid-connected new energy is improved.

Description

Technical Field

[0001] This invention belongs to the field of electricity metering technology, specifically, it relates to an electricity metering method and device for grid-connected new energy sources. Background Technology

[0002] With the rapid development of new energy sources, a large number of new energy power generation systems, mainly photovoltaic, are being connected to the power grid. The accuracy of grid metering after new energy is connected to the grid directly affects the fairness of transactions among power generation, transmission and distribution, and users. Therefore, accurate metering of electricity from new energy grid connections is receiving increasing attention.

[0003] In current electricity metering for grid-connected renewable energy sources, the presence of harmonics can cause meters to overestimate actual energy consumption, leading to significant metering errors. Therefore, it's necessary to determine error compensation based on harmonic information. Current technology typically involves personnel performing curve fitting on voltage and current data to determine harmonic characteristics and calculate error compensation coefficients. However, this method heavily relies on the expertise of personnel, making accurate determination of harmonic characteristics and error compensation coefficients difficult, and also incurring excessive labor costs. Furthermore, current error compensation calculations often lack consideration for the impact of interharmonics and non-sinusoidal signals on metering accuracy. This lack of consideration significantly affects the meter's accuracy, resulting in unreliable error compensation coefficients and leading to the continuous accumulation of metering errors. In error compensation calculation, time-varying error analysis of electricity meters is also a very important part. Currently, time-varying error analysis is mainly carried out through state-space equations. However, this method is still not accurate enough for time-varying error analysis, which can easily lead to a large deviation between the obtained time-varying error data and the actual situation, so that the accuracy of electricity metering for new energy grid connection fails to meet the target. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a method and device for metering electricity in grid-connected new energy sources. This method and device accurately compensate for errors caused by harmonic and non-sinusoidal signals, thereby improving the accuracy of electricity metering and achieving a more ideal effect for electricity metering in grid-connected new energy sources.

[0005] The present invention adopts the following technical solution.

[0006] This invention proposes a method for metering electricity consumption when connecting new energy sources to the grid, comprising:

[0007] The grid topology, grid voltage, and grid current under the new energy grid-connected operating conditions are obtained and preprocessed; among them, the new energy grid-connected operating conditions include power flow transformation operating conditions, low power factor operating conditions, bidirectional wide range operating conditions, and wide bandwidth operating conditions.

[0008] Based on the preprocessed grid voltage and current, harmonic time-frequency domain analysis data and harmonic current frequency distribution data are obtained using time-frequency domain analysis. A causal directed acyclic graph (DAG) is constructed based on the preprocessed grid topology. The harmonic influence intensity is determined based on the harmonic time-frequency domain analysis data and harmonic current frequency distribution data, using the DAG. The harmonic time-frequency domain analysis data, harmonic current frequency distribution data, and harmonic influence intensity are integrated to obtain broadband harmonic characteristic information.

[0009] Phase rotation is performed on the spectral coefficients corresponding to the preprocessed grid voltage and grid current. Based on the phase-rotated spectral coefficients, the first frequency, first amplitude, and first phase of the interharmonic are calculated. Wavelet analysis is used to calculate the second frequency, second amplitude, and second phase of the interharmonic from the preprocessed grid voltage and grid current. The average values ​​of the first and second frequencies, the first and second amplitudes, and the first and second phases of the interharmonic are used as the interharmonic characteristic information.

[0010] The influence of broadband harmonics on electricity metering errors is used to generate broadband harmonic characteristic vectors and first distortion characteristic vectors of voltage and current to construct broadband harmonic influence factors. The influence of interharmonics on electricity metering errors is used to generate interharmonic characteristic vectors and second distortion characteristic vectors of voltage and current to construct interharmonic influence factors. The influence factors, interharmonic influence factors, connection strengths between broadband harmonic influence factors, and connection strengths between interharmonic influence factors are used to construct an influence analysis topology network. After the broadband harmonic characteristic information and interharmonic characteristic information are input into the influence analysis topology network, the first error compensation parameter is calculated using the electricity metering error output by the influence analysis topology network.

[0011] A non-sinusoidal signal simulation model is established using the preprocessed power grid topology. Based on the non-sinusoidal signal simulation model, a non-sinusoidal signal is obtained by combining the preprocessed power grid voltage and current with a stepped wave signal. Error compensation calculations are performed on the non-sinusoidal signal to obtain the second error compensation parameter.

[0012] An operational trajectory matrix is ​​constructed using historical operational data of the electricity meter. Based on the operational trajectory matrix, time-varying error analysis of the electricity meter under the condition of new energy grid connection is performed to obtain the time-varying error data of the electricity meter.

[0013] The target error compensation parameter is calculated based on the first error compensation parameter, the second error compensation parameter, and the time-varying error data of the electricity meter. The electricity meter then performs electricity metering based on the target error compensation parameter to obtain the electricity metering result.

[0014] Preferably, based on the preprocessed grid voltage and grid current, harmonic time-frequency domain analysis data and harmonic current frequency distribution data are obtained using time-frequency domain analysis, including:

[0015] Construct a power grid network topology model based on the preprocessed power grid topology;

[0016] Determine the frequency domain analysis window; based on the preprocessed grid voltage and grid current, perform a fast Fourier transform using the frequency domain analysis window to obtain the target spectrum, and calculate the frequency, amplitude, and phase of the harmonics in the target spectrum;

[0017] Harmonic time-frequency domain analysis is performed based on the frequency, amplitude, and phase of the harmonics to obtain harmonic time-frequency domain analysis data; harmonic current frequency distribution analysis is performed based on the harmonic time-frequency domain analysis data to obtain harmonic current frequency distribution data.

[0018] Preferably, the power grid network topology model includes nodes and directed edges. The nodes include new energy grid connection points, load nodes, and other nodes. The directed edges are the transmission lines connecting each node, with the power flow direction on the transmission lines serving as the direction of each directed edge.

[0019] Preferably, a stable frequency reference point and a frequency domain range are set to determine the frequency domain analysis window, wherein the stable frequency reference point is the rated frequency of the power grid, and the frequency domain range is symmetrical and meets the requirements of power industry standards.

[0020] Preferably, the harmonic time series is embedded with time delay according to the frequency, amplitude and phase of the harmonic to obtain the phase space trajectory. Geometric analysis is performed on the phase space trajectory to obtain the Lyapunov characteristic index. Time-frequency domain analysis is performed based on the Lyapunov characteristic index and empirical mode decomposition to obtain the time-frequency trajectory and energy time-frequency distribution of the harmonic signal, which serve as the harmonic time-frequency domain analysis data.

[0021] Harmonic current frequency distribution analysis is performed based on harmonic time-frequency domain analysis data. The frequency of harmonic current time sequence is separated according to the harmonic time-frequency domain analysis data to obtain the harmonic current frequency, harmonic current frequency content rate and frequency duration, which are used as harmonic current frequency distribution data.

[0022] Preferably, a causal directed acyclic graph (DAG) is constructed based on the power grid network topology model; based on harmonic time-frequency domain analysis data and harmonic current frequency distribution data, harmonic influence intensity analysis is performed on the causal DAG to obtain the harmonic influence intensity, including:

[0023] The causal relationship discovery algorithm learns the nodes and directed edges in the power grid network topology model to obtain the causal strength between nodes.

[0024] The causal strength matrix is ​​constructed by combining the causal strength between nodes based on the Granger causality test, and then normalized to obtain the normalized causal strength matrix.

[0025] By connecting the nodes and directed edges of the directed acyclic graph based on the normalized causal strength matrix, a causal directed acyclic graph is obtained.

[0026] Based on harmonic time-frequency domain analysis data and harmonic current frequency distribution data, the harmonic amplitude time series and harmonic energy of each node in the causal directed acyclic graph are calculated. The harmonic contribution value is calculated based on the harmonic amplitude time series and harmonic energy of each node. The difference between the harmonic contribution value and the harmonic limit is taken as the harmonic influence intensity. The harmonic limit is set according to national standards.

[0027] Preferably, the preprocessed grid voltage and grid current are subjected to discrete Fourier transform to obtain spectral coefficients; the spectral coefficients are then phase-rotated, and the first frequency, first amplitude, and first phase of the interharmonic are calculated based on the phase-rotated spectral coefficients.

[0028] The preprocessed grid voltage and grid current are converted into analog voltage and analog current signals. Wavelet packet decomposition is performed on the analog voltage and analog current signals using wavelet basis functions to obtain frequency band wavelet coefficients. The reconstructed signal is obtained based on the frequency band wavelet coefficients. The target wavelet basis function is determined using the reconstructed signal, the analog voltage signal, and the analog current signal. The interharmonic components of the preprocessed grid voltage and grid current are detected using the target wavelet basis function to obtain the second frequency, second phase, and second amplitude of the corresponding interharmonics.

[0029] The average of the first and second frequencies of the interharmonic is calculated as the target frequency of the interharmonic. The average of the first and second amplitudes of the interharmonic is calculated as the target amplitude of the interharmonic. The average of the first and second phases of the interharmonic is calculated as the target phase of the interharmonic. The target frequency, target amplitude, and target phase of the interharmonic are used as the characteristic information of the interharmonic.

[0030] Preferably, the number of wavelet packet decomposition layers is determined based on the sampling frequency and fundamental frequency of the analog voltage and current signals, satisfying the following relationship:

[0031]

[0032] In the formula, n is the number of wavelet packet decomposition layers, p is the sampling frequency, j is the fundamental frequency, and σ is the correction coefficient.

[0033] Preferably, the covariance matrix of the error dataset of broadband harmonics affecting electricity metering is obtained as the first feature data matrix, and the covariance matrix of the error dataset of interharmonics affecting electricity metering is obtained as the second feature matrix; matrix element composition analysis is performed on the first feature data matrix to generate broadband harmonic feature vectors and first distortion feature vectors of voltage and current; matrix element composition analysis is performed on the second feature matrix to generate interharmonic feature vectors and second distortion feature vectors of voltage and current.

[0034] The broadband harmonic eigenvector and the first distortion eigenvector of voltage and current are subjected to feature interaction to determine the influence intensity of broadband harmonics and the influence intensity of broadband harmonics on power metering errors, which are used as broadband harmonic influence factors. The interharmonic eigenvector and the second distortion eigenvector of voltage and current are subjected to interharmonic influence factor analysis to determine the influence intensity of harmonics and the influence intensity of interharmonics on power metering errors, which are used as interharmonic influence factors.

[0035] Determine the first connection strength between each broadband harmonic influence factor and the second connection strength between each interharmonic influence factor. Construct a first analysis network based on the first connection strength and the broadband harmonic influence factors, and construct a second analysis network based on the second connection strength and the interharmonic influence factors. The first and second analysis networks constitute the influence analysis topology network.

[0036] After the broadband harmonic characteristic information and interharmonic characteristic information are input into the influence analysis topology network, the first error compensation parameter is calculated using the first error compensation function for the power metering error output by the influence analysis topology network; wherein, the first error compensation function is a preset compensation function.

[0037] Preferably, a non-sinusoidal signal simulation model is established using the preprocessed power grid topology; based on the non-sinusoidal signal simulation model, Fourier transform calculation is performed using the preprocessed power grid voltage and current combined with the stepped wave signal to obtain the Fourier transform coefficients; based on the Fourier transform coefficients, the active power measurement algorithm and reactive power measurement algorithm are used to calculate the non-sinusoidal signal data to obtain the non-sinusoidal signal data.

[0038] Based on non-sinusoidal signal data, error compensation is calculated using the second error compensation function to obtain the second error compensation parameters.

[0039] Preferably, historical operating data of the electricity meter is acquired and subjected to time-series chaotic analysis to obtain time-series chaotic data;

[0040] Establish a Markov chain, and construct the running trajectory matrix based on the Markov chain using temporal chaotic data combined with a bidirectional Bayesian network;

[0041] Based on the operation trajectory matrix combined with the preset environmental change relationship table and the preset load change relationship table, the time-varying error analysis of the electricity meter under the grid connection condition of new energy is carried out to obtain the time-varying error data of the electricity meter.

[0042] Preferably, the target error compensation parameter is the weighted sum of the first error compensation parameter, the second error compensation parameter, and the time-varying error data of the electricity meter.

[0043] The calibration parameters of the electricity meter are determined based on the target error compensation parameters. The electricity meter is then calibrated based on these parameters to obtain the calibrated electricity meter. Finally, the calibrated electricity meter is used to measure electricity based on the target error compensation parameters to obtain the electricity metering result.

[0044] This invention also proposes a power metering device for grid-connected renewable energy sources, comprising:

[0045] The data preprocessing module is used to acquire and preprocess the grid topology, grid voltage, and grid current under the new energy grid-connected operating conditions. Among them, the new energy grid-connected operating conditions include power flow transformation operating conditions, low power factor operating conditions, bidirectional wide range operating conditions, and wide bandwidth operating conditions.

[0046] The broadband harmonic feature extraction module is used to obtain harmonic time-frequency domain analysis data and harmonic current frequency distribution data based on the preprocessed grid voltage and grid current using time-frequency domain analysis; construct a causal directed acyclic graph based on the preprocessed grid topology; determine the harmonic influence intensity based on the causal directed acyclic graph using the harmonic time-frequency domain analysis data and harmonic current frequency distribution data; and integrate the harmonic time-frequency domain analysis data, harmonic current frequency distribution data, and harmonic influence intensity to obtain broadband harmonic feature information.

[0047] The interharmonic feature extraction module is used to perform phase rotation on the spectral coefficients corresponding to the preprocessed grid voltage and grid current, and calculate the first frequency, first amplitude, and first phase of the interharmonic based on the phase-rotated spectral coefficients; it also uses wavelet analysis to calculate the second frequency, second amplitude, and second phase of the interharmonic based on the preprocessed grid voltage and grid current; and uses the average of the first and second frequencies, the average of the first and second amplitudes, and the average of the first and second phases as the interharmonic feature information.

[0048] The first error compensation parameter generation module is used to generate broadband harmonic characteristic vectors and first distortion characteristic vectors of voltage and current based on the error caused by broadband harmonics affecting power metering, in order to construct broadband harmonic influence factors; it also generates interharmonic characteristic vectors and second distortion characteristic vectors of voltage and current based on the error caused by interharmonics affecting power metering, in order to construct interharmonic influence factors; it constructs an influence analysis topology network using the broadband harmonic influence factors, interharmonic influence factors, the connection strength between each broadband harmonic influence factor, and the connection strength between each interharmonic influence factor; after the broadband harmonic characteristic information and interharmonic characteristic information are input into the influence analysis topology network, the first error compensation parameter is calculated using the power metering error output by the influence analysis topology network.

[0049] The second error compensation parameter generation module is used to establish a non-sinusoidal signal simulation model using the preprocessed power grid topology; based on the non-sinusoidal signal simulation model, it uses the preprocessed power grid voltage and current combined with the stepped wave signal to obtain a non-sinusoidal signal; and performs error compensation calculation on the non-sinusoidal signal to obtain the second error compensation parameter.

[0050] The electricity meter time-varying error analysis module is used to construct an operating trajectory matrix using the historical operating data of the electricity meter, and to perform time-varying error analysis of the electricity meter under the new energy grid connection condition based on the operating trajectory matrix, so as to obtain the time-varying error data of the electricity meter.

[0051] The electricity metering module is used to calculate the target error compensation parameter based on the first error compensation parameter, the second error compensation parameter and the time-varying error data of the electricity meter. The electricity meter performs electricity metering based on the target error compensation parameter to obtain the electricity metering result.

[0052] A terminal includes a processor and a storage medium; the storage medium is used to store instructions; the processor is used to perform operations according to the instructions to execute the steps of a method.

[0053] A computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the steps of a method.

[0054] The beneficial effects of this invention are as follows: compared with the prior art, it at least includes the following: based on preprocessed power grid topology data, voltage waveform data, and current waveform data, it uses causal inference analysis to extract broadband harmonic features, achieving more refined broadband harmonic feature extraction and making the obtained broadband harmonic feature information more comprehensive and specific. Based on the influence analysis topology network, it generates a first error compensation parameter using broadband harmonic feature information and interharmonic feature information, and generates a second error compensation parameter based on the non-sinusoidal signal data. By incorporating interharmonic feature information and non-sinusoidal signal data into the generation of error compensation parameters, the generated error compensation parameters become more accurate. Based on the operating trajectory matrix, it performs time-varying error analysis of energy meters under complex operating conditions, improving the accuracy of time-varying error analysis of energy meters and avoiding excessive deviation between the obtained time-varying error data of energy meters and the actual situation. By calculating the target error compensation parameter using the first error compensation parameter, the second error compensation parameter, and the time-varying error data of the electricity meter, the electricity meter can be used for electricity metering. This avoids the continuous accumulation of electricity metering errors, greatly improves the accuracy of electricity metering, and enables the electricity metering of new energy grid-connected devices to achieve a more ideal effect. Attached Figure Description

[0055] Figure 1 This is a flowchart illustrating the electricity metering method for grid-connected new energy sources proposed in this invention.

[0056] Figure 2 This is a schematic diagram of the structural composition of the power metering device for grid connection of new energy sources in an embodiment of the present invention. Detailed Implementation

[0057] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of this invention. The embodiments described in this application are merely some embodiments of this invention, and not all embodiments. Based on the spirit of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of this invention.

[0058] This invention provides a method for metering electricity in grid-connected renewable energy sources, such as... Figure 1 As shown, it includes:

[0059] Step 1: Obtain the grid topology, grid voltage, and grid current under the new energy grid connection conditions and perform preprocessing; among which, the new energy grid connection conditions include power flow transformation conditions, low power factor conditions, bidirectional wide range conditions, and wide bandwidth conditions.

[0060] Specifically, the topology data, voltage waveform data, and current waveform data of the power grid under the condition of new energy grid connection are acquired, and the topology data, voltage waveform data, and current waveform data are preprocessed to obtain preprocessed topology data, voltage waveform data, and current waveform data.

[0061] In the specific implementation of this invention, after the new energy source is connected to the grid, a corresponding data acquisition unit is deployed to acquire grid topology data, voltage waveform data, and current waveform data under complex operating conditions. The grid topology data includes the connection relationships of lines, switches, and transformers. The voltage waveform data represents voltage changes over time, and the current waveform data represents current changes over time. Complex operating conditions include power flow transformation, low power factor, bidirectional wide range, and wide bandwidth. Acquiring data under complex operating conditions is beneficial for analyzing error compensation in electricity metering under these complex conditions. The grid topology data, voltage waveform data, and current waveform data are preprocessed, including data cleaning, data repair, abnormal data removal, and missing value imputation, to obtain preprocessed grid topology data, voltage waveform data, and current waveform data.

[0062] Step 2: Based on the preprocessed grid voltage and grid current, time-frequency domain analysis is used to obtain harmonic time-frequency domain analysis data and harmonic current frequency distribution data; based on the preprocessed grid topology, a causal directed acyclic graph is constructed; based on the harmonic time-frequency domain analysis data and harmonic current frequency distribution data, the harmonic influence intensity is determined based on the causal directed acyclic graph; the harmonic time-frequency domain analysis data, harmonic current frequency distribution data, and harmonic influence intensity are integrated to obtain broadband harmonic characteristic information.

[0063] Specifically, step 2 includes:

[0064] Step 2.1: Construct a power grid network topology model based on the preprocessed power grid topology;

[0065] In this embodiment, the power grid network topology model includes nodes and directed edges. The nodes include new energy grid connection points, load nodes, and other nodes. The directed edges are transmission lines connecting each node, with the power flow direction on the transmission lines serving as the direction of each directed edge.

[0066] Step 2.2, determine the frequency domain analysis window; based on the preprocessed grid voltage and grid current, use the frequency domain analysis window to perform a fast Fourier transform to obtain the target spectrum, and calculate the frequency, amplitude and phase of the harmonics in the target spectrum;

[0067] In this embodiment, a stable frequency reference point and a frequency domain range are set to determine the frequency domain analysis window. The stable frequency reference point is the rated frequency of the power grid, and the frequency domain range is symmetrical and meets the requirements of power industry standards.

[0068] The voltage and current signals are converted from the time domain to the frequency domain by using Fast Fourier Transform to reveal their spectra, obtain the target spectrum, and then calculate the frequency, amplitude, and phase of the harmonics in the target spectrum.

[0069] Step 2.3: Perform harmonic time-frequency domain analysis based on the frequency, amplitude, and phase of the harmonics to obtain harmonic time-frequency domain analysis data; perform harmonic current frequency distribution analysis based on the harmonic time-frequency domain analysis data to obtain harmonic current frequency distribution data.

[0070] In this embodiment, the harmonic time series is time-delayed and embedded according to the frequency, amplitude, and phase of the harmonics to obtain the phase space trajectory. Geometric analysis is performed on the phase space trajectory to obtain the Lyapunov characteristic index. Time-frequency domain analysis is then performed based on the Lyapunov characteristic index and empirical mode decomposition to obtain the time-frequency trajectory and energy time-frequency distribution of the harmonic signal, serving as harmonic time-frequency domain analysis data. Based on the harmonic time-frequency domain analysis data, harmonic current frequency distribution analysis is performed. Frequency separation of the harmonic current time series is then performed based on the harmonic time-frequency domain analysis data to obtain the harmonic current frequency, harmonic current frequency content, and frequency duration, serving as harmonic current frequency distribution data.

[0071] Step 2.4: Based on the power grid network topology model, construct a causal directed acyclic graph; based on the harmonic time-frequency domain analysis data and harmonic current frequency distribution data, perform harmonic influence intensity analysis based on the causal directed acyclic graph to obtain the harmonic influence intensity.

[0072] include:

[0073] Step 2.4.1: Based on the causal relationship discovery algorithm, learn the nodes and directed edges in the power grid network topology model to obtain the causal strength between nodes;

[0074] Step 2.4.2: Based on the Granger causality test, combine the causal strength between nodes to construct a causal strength matrix, and normalize the causal strength matrix to obtain a normalized causal strength matrix.

[0075] Step 2.4.3: Connect the nodes and directed edges of the directed acyclic graph according to the normalized causal strength matrix to obtain the causal directed acyclic graph.

[0076] In this invention, the power grid network topology model constructed based on the preprocessed power grid topology is a directed acyclic graph (DAG). The relationship between two nodes on the same edge is one of association, not causation. Therefore, a causal DAG needs to be constructed based on the power grid network topology model to eliminate interference from non-causal nodes. Based on harmonic time-frequency domain analysis data and harmonic current frequency distribution data, and using the causal DAG, a more accurate harmonic influence intensity can be determined, thus eliminating the impact of other harmonic interference sources on the metering of new energy grid connection.

[0077] Step 2.4.4: Based on the harmonic time-frequency domain analysis data and harmonic current frequency distribution data, calculate the harmonic amplitude time series and harmonic energy of each node in the causal directed acyclic graph; calculate the harmonic contribution value based on the harmonic amplitude time series and harmonic energy of each node, and use the difference between the harmonic contribution value and the harmonic limit as the harmonic influence intensity, wherein the harmonic limit is set according to national standards.

[0078] Step 2.5: Using a broadband harmonic feature integration model based on deep neural networks, the harmonic time-frequency domain analysis data, harmonic current frequency distribution data, and harmonic influence intensity are integrated to obtain broadband harmonic feature information.

[0079] Wideband harmonics refer to harmonics with wide frequency domain characteristics. In power systems, wideband harmonics mainly originate from new energy power generation and various types of loads. These factors lead to a continuous increase in the harmonic content of the power grid and a widening of the harmonic frequency domain, exhibiting wideband characteristics. Harmonic time-frequency domain analysis data, harmonic current frequency distribution data, and harmonic influence intensity data are input into a wideband harmonic feature integration model for feature integration. This model is a convergent model obtained by inputting sample datasets into a deep neural network for training, thus obtaining wideband harmonic feature information. This enables more refined wideband harmonic feature extraction, making the obtained wideband harmonic feature information more comprehensive and specific, and significantly reducing the complexity of manual feature engineering.

[0080] Step 3: Perform phase rotation on the spectral coefficients corresponding to the preprocessed grid voltage and grid current, and calculate the first frequency, first amplitude, and first phase of the interharmonic based on the phase-rotated spectral coefficients; use wavelet analysis to calculate the second frequency, second amplitude, and second phase of the interharmonic based on the preprocessed grid voltage and grid current; use the average of the first and second frequencies, the average of the first and second amplitudes, and the average of the first and second phases of the interharmonic as the interharmonic characteristic information.

[0081] Specifically, step 3 includes:

[0082] Step 3.1: Perform Discrete Fourier Transform on the preprocessed grid voltage and grid current to obtain the spectral coefficients; perform phase rotation on the spectral coefficients, and calculate the first frequency, first amplitude, and first phase of the interharmonic based on the phase-rotated spectral coefficients;

[0083] In this embodiment, the preprocessed grid voltage and grid current are subjected to Discrete Fourier Transform using a rectangular window with a preset fundamental frequency length. The Discrete Fourier Transform converts the data from the time domain to the frequency domain to obtain the spectral coefficients.

[0084] In the embodiment, since both the positive and negative frequency components of the interharmonic spectrum contain phase factors, and the phase factors affect the phase of the spectral coefficients as the spectral line position changes, phase rotation of the spectral coefficients can eliminate the influence of interharmonic spectrum leakage, achieve accurate calculation of interharmonic components, obtain the phase-rotated spectral coefficients, and perform spectral analysis on the phase-rotated spectral coefficients to obtain the first frequency, first amplitude, and first phase of the interharmonic.

[0085] Step 3.2: Convert the preprocessed grid voltage and grid current into analog voltage and analog current signals. Perform wavelet packet decomposition on the analog voltage and analog current signals using wavelet basis functions to obtain frequency band wavelet coefficients. Obtain the reconstructed signal based on the frequency band wavelet coefficients. Determine the target wavelet basis function using the reconstructed signal, analog voltage signal, and analog current signal. Detect interharmonic components on the preprocessed grid voltage and grid current using the target wavelet basis function to obtain the second frequency, second phase, and second amplitude of the corresponding interharmonics.

[0086] In this embodiment, the preprocessed grid voltage and grid current are converted into analog voltage and analog current signals. Wavelet packet decomposition is then performed on the analog voltage and analog current signals according to the wavelet packet decomposition level, yielding several corresponding frequency band wavelet coefficients. The wavelet packet decomposition level is determined based on the sampling frequency and fundamental frequency of the analog voltage and analog current signals, satisfying the following relationship:

[0087]

[0088] In the formula, n is the number of wavelet packet decomposition layers, p is the sampling frequency, j is the fundamental frequency, and σ is the correction coefficient;

[0089] In this embodiment, single-branch reconstruction is performed based on the frequency band wavelet coefficients to obtain several reconstructed signals. Fast Fourier Transform is performed on the reconstructed signals, analog voltage signals, and analog current signals to obtain the corresponding amplitude and phase information. Phase error and amplitude error are calculated based on the corresponding amplitude and phase information. Target wavelet basis function is determined based on phase error and amplitude error. Interharmonic components are detected on the preprocessed grid voltage and grid current using the target wavelet basis function to obtain the second frequency, second phase, and second amplitude of the corresponding interharmonics.

[0090] Step 3.3: Calculate the average of the first frequency and the second frequency of the interharmonic as the target frequency of the interharmonic, calculate the average of the first amplitude and the second amplitude of the interharmonic as the target amplitude of the interharmonic, and calculate the average of the first phase and the second phase of the interharmonic as the target phase of the interharmonic; the target frequency, target amplitude and target phase of the interharmonic are used as the characteristic information of the interharmonic.

[0091] This invention enables the accurate extraction of interharmonic characteristic information.

[0092] Step 4: Utilize the error caused by broadband harmonics affecting electricity metering to generate broadband harmonic characteristic vectors and the first distortion characteristic vectors of voltage and current, thus constructing a broadband harmonic influence factor; utilize the error caused by interharmonics affecting electricity metering to generate interharmonic characteristic vectors and the second distortion characteristic vectors of voltage and current, thus constructing an interharmonic influence factor; use the broadband harmonic influence factor, the interharmonic influence factor, the connection strength between each broadband harmonic influence factor, and the connection strength between each interharmonic influence factor to construct an influence analysis topology network; after inputting the broadband harmonic characteristic information and the interharmonic characteristic information into the influence analysis topology network, calculate the first error compensation parameter using the electricity metering error output by the influence analysis topology network.

[0093] Specifically, step 4 includes:

[0094] Step 4.1: Obtain the covariance matrix of the error dataset of broadband harmonics affecting electricity metering as the first feature data matrix, and obtain the covariance matrix of the error dataset of interharmonics affecting electricity metering as the second feature matrix; perform matrix element composition analysis on the first feature data matrix to generate broadband harmonic feature vectors and first distortion feature vectors of voltage and current; perform matrix element composition analysis on the second feature matrix to generate interharmonic feature vectors and second distortion feature vectors of voltage and current.

[0095] In this embodiment, error datasets of the impact of broadband harmonics on electricity metering and error datasets of the impact of interharmonics on electricity metering are obtained from the database. The error data of broadband harmonics on electricity metering includes numerical values ​​corresponding to physical quantities such as the distortion effect of broadband harmonics on grid voltage and grid current. The error data of interharmonics on electricity metering includes numerical values ​​corresponding to physical quantities such as the distortion effect of broadband harmonics on grid voltage and grid current. A first feature data matrix is ​​constructed using the numericalized error dataset of broadband harmonics on electricity metering, and a second feature matrix is ​​constructed using the numericalized error dataset of interharmonics on electricity metering.

[0096] The first feature data matrix is ​​constructed using the covariance matrix of the error dataset of electricity metering affected by broadband harmonics, and the second feature matrix is ​​constructed using the covariance matrix of the error dataset of electricity metering affected by interharmonics.

[0097] Step 4.2: Perform feature interaction on the broadband harmonic feature vector and the first distortion feature vector of voltage and current to determine the influence intensity of broadband harmonics and the influence intensity of broadband harmonics on power metering error, which are used as broadband harmonic influence factors; perform interharmonic influence factor analysis on the interharmonic feature vector and the second distortion feature vector of voltage and current to determine the influence intensity of harmonics and the influence intensity of interharmonics on power metering error, which are used as interharmonic influence factors.

[0098] Step 4.3: Determine the first connection strength between each broadband harmonic influence factor and the second connection strength between each interharmonic influence factor. Construct a first analysis network based on the first connection strength and the broadband harmonic influence factors, and construct a second analysis network based on the second connection strength and the interharmonic influence factors. The first analysis network and the second analysis network constitute the influence analysis topology network.

[0099] Step 4.4: After inputting the broadband harmonic characteristic information and interharmonic characteristic information into the influence analysis topology network, the first error compensation parameter is calculated using the first error compensation function for the power metering error output by the influence analysis topology network; wherein, the first error compensation function is a preset compensation function.

[0100] The first error compensation parameter obtained by this invention takes into account the influence of broadband harmonics and interharmonics on the power metering error, making the obtained error compensation parameter more reliable.

[0101] Step 5: Establish a non-sinusoidal signal simulation model using the preprocessed power grid topology; based on the non-sinusoidal signal simulation model, obtain a non-sinusoidal signal by combining the preprocessed power grid voltage and current with a stepped wave signal; perform error compensation calculation on the non-sinusoidal signal to obtain the second error compensation parameter.

[0102] Specifically, step 5 includes:

[0103] Step 5.1: Establish a non-sinusoidal signal simulation model using the preprocessed power grid topology; Based on the non-sinusoidal signal simulation model, perform Fourier transform calculations using the preprocessed power grid voltage and current combined with the stepped wave signal to obtain the Fourier transform coefficients; Based on the Fourier transform coefficients, use the active power measurement algorithm and reactive power measurement algorithm to calculate the non-sinusoidal signal data to obtain the non-sinusoidal signal data.

[0104] In this embodiment, the simulation platform software constructs a non-sinusoidal signal simulation model based on the topology data of the new energy grid connection. Under the condition that the voltage and current harmonic model of the non-sinusoidal signal is the fundamental wave, the effective values ​​of voltage and current of harmonics below the preset order and the phase angle of voltage and current of each harmonic are selected to construct the non-sinusoidal signal simulation model.

[0105] A stepped wave signal is a non-sinusoidal waveform signal characterized by maintaining a constant amplitude within a specified time period and then jumping to another constant amplitude value within a short period of time. In a non-sinusoidal signal simulation model, the stepped wave signal is sampled differentially with a preset non-sinusoidal signal at equal intervals to obtain the difference value. Based on the difference value and the preprocessed grid voltage and grid current, a discrete Fourier transform is performed at the preset sampling points to obtain the Fourier transform coefficients.

[0106] In this embodiment, the active power measurement algorithm adopts the frequency domain angle active power algorithm, and the reactive power measurement algorithm adopts the root mean square algorithm. Then, the simulation calculation of non-sinusoidal signal data is performed to obtain non-sinusoidal signal data.

[0107] Step 5.2: Based on the non-sinusoidal signal data, perform error compensation calculation using the second error compensation function to obtain the second error compensation parameters.

[0108] In this embodiment, a non-sinusoidal signal refers to a signal whose waveform does not conform to the characteristics of a sine function. The presence of a non-sinusoidal signal can cause fluctuations in grid voltage, affecting the metering performance of the electricity meter. Its stacking with broadband harmonics and interharmonics can lead to the continuous accumulation of grid metering errors. Therefore, taking into account the impact of non-sinusoidal signals on electricity metering can prevent the continuous accumulation of grid metering errors.

[0109] The second error compensation function is a preset compensation function.

[0110] Step 6: Construct an operation trajectory matrix using the historical operation data of the electricity meter, and perform time-varying error analysis of the electricity meter under the new energy grid connection condition based on the operation trajectory matrix to obtain the time-varying error data of the electricity meter.

[0111] Specifically, step 6 includes:

[0112] Step 6.1: Obtain historical operating data of the electricity meter and perform time-series chaotic analysis to obtain time-series chaotic data;

[0113] In the implementation example, historical operating information of electricity meters was retrieved from the database. This information includes operating time, frequency of anomalies in each time period, maintenance records, and load conditions. Temporal chaotic analysis was then performed on this historical operating information. Temporal chaotic analysis, also known as time series chaos analysis, is used to analyze nonlinear time series with chaotic characteristics. The Lyapunov exponent was used to perform temporal chaotic analysis on the historical operating information of the electricity meters to obtain the chaotic characteristics of the time series within the historical operating information, thus acquiring temporal chaotic data.

[0114] Step 6.2: Establish a Markov chain, and construct the running trajectory matrix based on the Markov chain using time-series chaotic data combined with a bidirectional Bayesian network;

[0115] In this embodiment, a Markov chain is established, and the frequency of anomalies is calculated using historical operating information of the electricity meter. The Markov chain is then constructed based on the anomaly frequency and the electricity meter's state transition stages. Based on the Markov chain, an operational trajectory matrix is ​​constructed using the temporal chaotic data and a bidirectional Bayesian network. A time series matrix is ​​formed based on the temporal chaotic data. The probability of anomalies is calculated using the Markov chain and the bidirectional Bayesian network, yielding the anomaly probability calculation result. This result is then combined with the time series matrix to obtain the operational trajectory matrix.

[0116] Step 6.3: Based on the operation trajectory matrix combined with the preset environmental change relationship table and the preset load change relationship table, perform time-varying error analysis of the electricity meter under the new energy grid connection condition to obtain the time-varying error data of the electricity meter.

[0117] In this embodiment, a preset environmental change relationship table represents the performance changes of the electricity meter under different environments, and a preset load change relationship table represents the performance changes of the electricity meter under different load conditions. By combining the running trajectory matrix with the current electricity meter running data, a loss analysis of the electricity meter over time is performed to obtain the corresponding loss data. Based on the current environmental data and current load data, combined with the preset environmental change relationship table and the preset load change relationship table, the performance data of the electricity meter over time is determined, thereby constructing the time-varying error data of the electricity meter. Time-varying error refers to the error caused by the change of certain parameters or characteristics of the equipment or system over time.

[0118] Step 7: Calculate the target error compensation parameter based on the first error compensation parameter, the second error compensation parameter, and the time-varying error data of the electricity meter. The electricity meter performs electricity metering based on the target error compensation parameter to obtain the electricity metering result.

[0119] Specifically, step 7 includes:

[0120] Step 7.1: Use the weighted sum of the first error compensation parameter, the second error compensation parameter, and the time-varying error data of the energy meter as the target error compensation parameter;

[0121] In this embodiment, different parameters correspond to different weighting coefficients, and the corresponding weighting coefficients can be matched in the database.

[0122] Step 7.2: Determine the energy meter calibration parameters based on the target error compensation parameters, calibrate the energy meter based on the energy meter calibration parameters, obtain the calibrated energy meter, and use the calibrated energy meter to measure energy based on the target error compensation parameters to obtain the energy measurement result.

[0123] In this embodiment, the power metering error is corrected by using a target error compensation parameter to minimize the power metering error and obtain the power metering result.

[0124] This invention also provides an energy metering device for grid-connected renewable energy sources, such as... Figure 2 As shown, it includes:

[0125] The data preprocessing module 31 is used to acquire and preprocess the grid topology, grid voltage and grid current under the new energy grid connection conditions; wherein, the new energy grid connection conditions include power flow transformation conditions, low power factor conditions, bidirectional wide range conditions and wide frequency band conditions.

[0126] The broadband harmonic feature extraction module 32 is used to obtain harmonic time-frequency domain analysis data and harmonic current frequency distribution data by using time-frequency domain analysis based on the preprocessed grid voltage and grid current; construct a causal directed acyclic graph based on the preprocessed grid topology; determine the harmonic influence intensity based on the causal directed acyclic graph using the harmonic time-frequency domain analysis data and harmonic current frequency distribution data; and integrate the harmonic time-frequency domain analysis data, harmonic current frequency distribution data, and harmonic influence intensity to obtain broadband harmonic feature information.

[0127] The interharmonic feature extraction module 33 is used to perform phase rotation on the spectral coefficients corresponding to the preprocessed grid voltage and grid current, and calculate the first frequency, first amplitude, and first phase of the interharmonic based on the phase-rotated spectral coefficients; it also uses wavelet analysis to calculate the second frequency, second amplitude, and second phase of the interharmonic based on the preprocessed grid voltage and grid current; and uses the average value of the first and second frequencies, the average value of the first and second amplitudes, and the average value of the first and second phases as the interharmonic feature information.

[0128] The first error compensation parameter generation module 34 is used to generate broadband harmonic feature vectors and first distortion feature vectors of voltage and current using the error caused by the influence of broadband harmonics on power metering, so as to construct broadband harmonic influence factors; to generate interharmonic feature vectors and second distortion feature vectors of voltage and current using the error caused by the influence of interharmonics on power metering, so as to construct interharmonic influence factors; to construct influence analysis topology network using broadband harmonic influence factors, interharmonic influence factors, connection strength between broadband harmonic influence factors and connection strength between interharmonic influence factors; after broadband harmonic feature information and interharmonic feature information are input into influence analysis topology network, the first error compensation parameter is calculated using the power metering error output by influence analysis topology network.

[0129] The second error compensation parameter generation module 35 is used to establish a non-sinusoidal signal simulation model using the preprocessed power grid topology; based on the non-sinusoidal signal simulation model, it uses the preprocessed power grid voltage and current combined with the stepped wave signal to obtain a non-sinusoidal signal; and performs error compensation calculation on the non-sinusoidal signal to obtain the second error compensation parameter.

[0130] The time-varying error analysis module 36 for electricity meters is used to construct an operating trajectory matrix using the historical operating data of the electricity meter, and to perform time-varying error analysis of the electricity meter under the new energy grid connection condition based on the operating trajectory matrix, so as to obtain the time-varying error data of the electricity meter.

[0131] The electricity metering module 37 is used to calculate the target error compensation parameter based on the first error compensation parameter, the second error compensation parameter and the time-varying error data of the electricity meter. The electricity meter performs electricity metering based on the target error compensation parameter to obtain the electricity metering result.

[0132] This disclosure can be a system, method, and / or computer program product. A computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for causing a processor to implement various aspects of this disclosure.

[0133] Computer-readable storage media can be tangible devices capable of holding and storing instructions for use by an instruction execution device. Computer-readable storage media can be, for example—but not limited to—electrical storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory sticks, floppy disks, mechanical encoding devices, such as punch cards or recessed protrusions storing instructions thereon, and any suitable combination of the foregoing. The computer-readable storage media used herein are not to be construed as transient signals themselves, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses through fiber optic cables), or electrical signals transmitted through wires.

[0134] The computer-readable program instructions described herein can be downloaded from computer-readable storage media to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to the computer-readable storage media in the respective computing / processing device.

[0135] Computer program instructions used to perform the operations of this disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, status setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The computer-readable program instructions may execute entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing the status information of the computer-readable program instructions to implement various aspects of this disclosure.

[0136] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A method for metering electricity connected to a new energy source, characterized in that, include: The grid topology, grid voltage, and grid current under the new energy grid-connected operating conditions are obtained and preprocessed; among them, the new energy grid-connected operating conditions include power flow transformation operating conditions, low power factor operating conditions, bidirectional wide range operating conditions, and wide bandwidth operating conditions. Based on the preprocessed grid voltage and current, harmonic time-frequency domain analysis data and harmonic current frequency distribution data are obtained using time-frequency domain analysis. A causal directed acyclic graph (DAG) is constructed based on the preprocessed grid topology. The harmonic influence intensity is determined based on the harmonic time-frequency domain analysis data and harmonic current frequency distribution data, using the DAG. The harmonic time-frequency domain analysis data, harmonic current frequency distribution data, and harmonic influence intensity are integrated to obtain broadband harmonic characteristic information. Phase rotation is performed on the spectral coefficients corresponding to the preprocessed grid voltage and grid current. Based on the phase-rotated spectral coefficients, the first frequency, first amplitude, and first phase of the interharmonic are calculated. Wavelet analysis is used to calculate the second frequency, second amplitude, and second phase of the interharmonic from the preprocessed grid voltage and grid current. The average values ​​of the first and second frequencies, the first and second amplitudes, and the first and second phases of the interharmonic are used as the interharmonic characteristic information. The influence of broadband harmonics on electricity metering errors is used to generate broadband harmonic characteristic vectors and first distortion characteristic vectors of voltage and current to construct broadband harmonic influence factors. The influence of interharmonics on electricity metering errors is used to generate interharmonic characteristic vectors and second distortion characteristic vectors of voltage and current to construct interharmonic influence factors. The influence factors, interharmonic influence factors, connection strengths between broadband harmonic influence factors, and connection strengths between interharmonic influence factors are used to construct an influence analysis topology network. After the broadband harmonic characteristic information and interharmonic characteristic information are input into the influence analysis topology network, the first error compensation parameter is calculated using the electricity metering error output by the influence analysis topology network. A non-sinusoidal signal simulation model is established using the preprocessed power grid topology. Based on the non-sinusoidal signal simulation model, a non-sinusoidal signal is obtained by combining the preprocessed power grid voltage and current with a stepped wave signal. Error compensation calculations are performed on the non-sinusoidal signal to obtain the second error compensation parameter. An operational trajectory matrix is ​​constructed using historical operational data of the electricity meter. Based on the operational trajectory matrix, time-varying error analysis of the electricity meter under the condition of new energy grid connection is performed to obtain the time-varying error data of the electricity meter. The target error compensation parameter is calculated based on the first error compensation parameter, the second error compensation parameter, and the time-varying error data of the electricity meter. The electricity meter then performs electricity metering based on the target error compensation parameter to obtain the electricity metering result.

2. The method for metering electricity in grid-connected new energy sources according to claim 1, characterized in that, include: Based on the preprocessed grid voltage and grid current, harmonic time-frequency domain analysis data and harmonic current frequency distribution data are obtained using time-frequency domain analysis, including: Construct a power grid network topology model based on the preprocessed power grid topology; Determine the frequency domain analysis window; based on the preprocessed grid voltage and grid current, perform a fast Fourier transform using the frequency domain analysis window to obtain the target spectrum, and calculate the frequency, amplitude, and phase of the harmonics in the target spectrum; Harmonic time-frequency domain analysis is performed based on the frequency, amplitude, and phase of the harmonics to obtain harmonic time-frequency domain analysis data; harmonic current frequency distribution analysis is performed based on the harmonic time-frequency domain analysis data to obtain harmonic current frequency distribution data.

3. The method for metering electricity in new energy grid connection according to claim 2, characterized in that, include: The power grid network topology model includes nodes and directed edges. Nodes include new energy grid connection points, load nodes, and other nodes. Directed edges are transmission lines that connect the nodes, with the power flow direction on the transmission lines serving as the direction of each directed edge.

4. The method for metering electricity in new energy grid connection according to claim 3, characterized in that, include: Set a stable frequency reference point and frequency domain range to determine the frequency domain analysis window. The stable frequency reference point is the rated frequency of the power grid, and the frequency domain range is symmetrical and meets the requirements of power industry standards.

5. The method for metering electricity in new energy grid connection according to claim 4, characterized in that, include: The harmonic time series is embedded with time delay based on the frequency, amplitude and phase of the harmonics to obtain the phase space trajectory. Geometric analysis is performed on the phase space trajectory to obtain the Lyapunov characteristic index. Based on the Lyapunov characteristic index and empirical mode decomposition, time-frequency domain analysis is performed to obtain the time-frequency trajectory and energy time-frequency distribution of the harmonic signal, which serve as the harmonic time-frequency domain analysis data. Harmonic current frequency distribution analysis is performed based on harmonic time-frequency domain analysis data. The frequency of harmonic current time sequence is separated according to the harmonic time-frequency domain analysis data to obtain the harmonic current frequency, harmonic current frequency content rate and frequency duration, which are used as harmonic current frequency distribution data.

6. The method for metering electricity in grid-connected new energy sources according to claim 3, characterized in that, include: Based on the power grid network topology model, a causal directed acyclic graph is constructed. Based on harmonic time-frequency domain analysis data and harmonic current frequency distribution data, harmonic influence intensity analysis is performed on a causal directed acyclic graph to obtain the harmonic influence intensity, including: learning the nodes and directed edges in the power grid network topology model based on the causal relationship discovery algorithm to obtain the causal intensity between nodes; The causal strength matrix is ​​constructed by combining the causal strength between nodes based on the Granger causality test, and then normalized to obtain the normalized causal strength matrix. By connecting the nodes and directed edges of the directed acyclic graph based on the normalized causal strength matrix, a causal directed acyclic graph is obtained. Based on harmonic time-frequency domain analysis data and harmonic current frequency distribution data, the harmonic amplitude time series and harmonic energy of each node in the causal directed acyclic graph are calculated. The harmonic contribution value is calculated based on the harmonic amplitude time series and harmonic energy of each node. The difference between the harmonic contribution value and the harmonic limit is taken as the harmonic influence intensity. The harmonic limit is set according to national standards.

7. The method for metering electricity in new energy grid connection according to claim 1, characterized in that, include: Discrete Fourier transform is performed on the preprocessed grid voltage and grid current to obtain the spectral coefficients; The spectral coefficients are phase-rotated, and the first frequency, first amplitude, and first phase of the interharmonic are calculated based on the phase-rotated spectral coefficients. The preprocessed grid voltage and grid current are converted into analog voltage and analog current signals. Wavelet packet decomposition is performed on the analog voltage and analog current signals using wavelet basis functions to obtain frequency band wavelet coefficients. The reconstructed signal is obtained based on the frequency band wavelet coefficients. The target wavelet basis function is determined using the reconstructed signal, the analog voltage signal, and the analog current signal. The target wavelet basis function is used to detect interharmonic components of the preprocessed grid voltage and grid current to obtain the second frequency, second phase and second amplitude of the corresponding interharmonics. The average of the first and second frequencies of the interharmonic is calculated as the target frequency of the interharmonic. The average of the first and second amplitudes of the interharmonic is calculated as the target amplitude of the interharmonic. The average of the first and second phases of the interharmonic is calculated as the target phase of the interharmonic. The target frequency, target amplitude, and target phase of the interharmonic are used as the characteristic information of the interharmonic.

8. The method for metering electricity in grid-connected new energy sources according to claim 7, characterized in that, include: The number of wavelet packet decomposition layers is determined based on the sampling frequency and fundamental frequency of the analog voltage and current signals, satisfying the following relationship: In the formula, The wavelet packet decomposition level is [number of layers]. Sampling frequency, For the fundamental frequency, This is the correction factor.

9. The method for metering electricity in new energy grid connection according to claim 1, characterized in that, include: The covariance matrix of the error dataset of broadband harmonics affecting electricity metering is obtained as the first feature data matrix, and the covariance matrix of the error dataset of interharmonics affecting electricity metering is obtained as the second feature matrix. Matrix element composition analysis is performed on the first feature data matrix to generate broadband harmonic feature vectors and first distortion feature vectors of voltage and current; matrix element composition analysis is performed on the second feature matrix to generate interharmonic feature vectors and second distortion feature vectors of voltage and current. The broadband harmonic eigenvector and the first distortion eigenvector of voltage and current are subjected to feature interaction to determine the influence intensity of broadband harmonics and the influence intensity of broadband harmonics on power metering errors, which are used as broadband harmonic influence factors. The interharmonic eigenvector and the second distortion eigenvector of voltage and current are subjected to interharmonic influence factor analysis to determine the influence intensity of harmonics and the influence intensity of interharmonics on power metering errors, which are used as interharmonic influence factors. Determine the first connection strength between each broadband harmonic influence factor and the second connection strength between each interharmonic influence factor. Construct a first analysis network based on the first connection strength and the broadband harmonic influence factors, and construct a second analysis network based on the second connection strength and the interharmonic influence factors. The first and second analysis networks constitute the influence analysis topology network. After the broadband harmonic characteristic information and interharmonic characteristic information are input into the influence analysis topology network, the first error compensation parameter is calculated using the first error compensation function for the power metering error output by the influence analysis topology network; wherein, the first error compensation function is a preset compensation function.

10. The method for metering electricity in grid-connected new energy sources according to claim 1, characterized in that, include: A non-sinusoidal signal simulation model is established using the preprocessed power grid topology. Based on the non-sinusoidal signal simulation model, Fourier transform calculations are performed using the preprocessed power grid voltage and current combined with the stepped wave signal to obtain the Fourier transform coefficients. Based on the Fourier transform coefficients, active power measurement algorithm and reactive power measurement algorithm are used to calculate non-sinusoidal signal data to obtain non-sinusoidal signal data; Based on non-sinusoidal signal data, error compensation is calculated using the second error compensation function to obtain the second error compensation parameters.

11. The method for metering electricity in grid-connected new energy sources according to claim 1, characterized in that, include: Historical operating data of electricity meters are acquired and subjected to time-series chaotic analysis to obtain time-series chaotic data; A Markov chain is established, and an operational trajectory matrix is ​​constructed based on the Markov chain using time-series chaotic data and a bidirectional Bayesian network. Based on the operational trajectory matrix, combined with a preset environmental change relationship table and a preset load change relationship table, time-varying error analysis of the electricity meter under the new energy grid-connected operating condition is performed to obtain the time-varying error data of the electricity meter.

12. The method for metering electricity in new energy grid connection according to claim 1, characterized in that, include: The target error compensation parameter is the weighted sum of the first error compensation parameter, the second error compensation parameter, and the time-varying error data of the energy meter. The calibration parameters of the electricity meter are determined based on the target error compensation parameters. The electricity meter is then calibrated based on these parameters to obtain the calibrated electricity meter. Finally, the calibrated electricity meter is used to measure electricity based on the target error compensation parameters to obtain the electricity metering result.

13. A power metering device for grid-connected new energy sources, characterized in that, include: The data preprocessing module is used to acquire and preprocess the grid topology, grid voltage, and grid current under the new energy grid-connected operating conditions. Among them, the new energy grid-connected operating conditions include power flow transformation operating conditions, low power factor operating conditions, bidirectional wide range operating conditions, and wide bandwidth operating conditions. The broadband harmonic feature extraction module is used to obtain harmonic time-frequency domain analysis data and harmonic current frequency distribution data based on the preprocessed grid voltage and grid current using time-frequency domain analysis; construct a causal directed acyclic graph based on the preprocessed grid topology; determine the harmonic influence intensity based on the causal directed acyclic graph using the harmonic time-frequency domain analysis data and harmonic current frequency distribution data; and integrate the harmonic time-frequency domain analysis data, harmonic current frequency distribution data, and harmonic influence intensity to obtain broadband harmonic feature information. The interharmonic feature extraction module is used to perform phase rotation on the spectral coefficients corresponding to the preprocessed grid voltage and grid current, and calculate the first frequency, first amplitude, and first phase of the interharmonic based on the phase-rotated spectral coefficients; it also uses wavelet analysis to calculate the second frequency, second amplitude, and second phase of the interharmonic based on the preprocessed grid voltage and grid current; and uses the average of the first and second frequencies, the average of the first and second amplitudes, and the average of the first and second phases as the interharmonic feature information. The first error compensation parameter generation module is used to generate broadband harmonic characteristic vectors and first distortion characteristic vectors of voltage and current based on the error caused by broadband harmonics affecting power metering, in order to construct broadband harmonic influence factors; it also generates interharmonic characteristic vectors and second distortion characteristic vectors of voltage and current based on the error caused by interharmonics affecting power metering, in order to construct interharmonic influence factors; it constructs an influence analysis topology network using the broadband harmonic influence factors, interharmonic influence factors, the connection strength between each broadband harmonic influence factor, and the connection strength between each interharmonic influence factor; after the broadband harmonic characteristic information and interharmonic characteristic information are input into the influence analysis topology network, the first error compensation parameter is calculated using the power metering error output by the influence analysis topology network. The second error compensation parameter generation module is used to establish a non-sinusoidal signal simulation model using the preprocessed power grid topology; based on the non-sinusoidal signal simulation model, it uses the preprocessed power grid voltage and current combined with the stepped wave signal to obtain a non-sinusoidal signal; and performs error compensation calculation on the non-sinusoidal signal to obtain the second error compensation parameter. The electricity meter time-varying error analysis module is used to construct an operating trajectory matrix using the historical operating data of the electricity meter, and to perform time-varying error analysis of the electricity meter under the new energy grid connection condition based on the operating trajectory matrix, so as to obtain the time-varying error data of the electricity meter. The electricity metering module is used to calculate the target error compensation parameter based on the first error compensation parameter, the second error compensation parameter and the time-varying error data of the electricity meter. The electricity meter performs electricity metering based on the target error compensation parameter to obtain the electricity metering result.

14. A terminal, comprising a processor and a storage medium; characterized in that: The storage medium is used to store instructions; The processor is configured to operate according to the instructions to perform the steps of the method according to any one of claims 1-12.

15. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the steps of the method according to any one of claims 1-12.

Images