Electric power meter self-calibration method, device, equipment, medium and product

Abstract

The invention discloses a self-calibration method, device, equipment, medium and product for an electric power meter, and relates to the field of calibration, and the method comprises the steps: obtaining a current power supply phase vector group in a set sampling period of an edge node according to the collected current power supply phase vector of each voltage channel in the electric power meter; vector difference offset processing is carried out on the current power supply phase vector set and the edge synchronization reference phase, and the comprehensive phase offset of the current period is obtained; after a power supply abnormal interruption event is detected, a zero point error estimation value is obtained according to the power supply phase vector extreme value set before power failure and the initial power supply phase vector set after power on; the zero error estimated value and the comprehensive phase offset are jointly input into the calibration correction function, the local correction factor set is updated, and a self-adaptive calibration parameter set is formed. The method is suitable for a remote deployment scene in a complex environment, and the long-term operation stability and the self-recovery capacity of the electric power meter are remarkably improved.

Description

Technical Field

[0001] This application relates to the field of calibration, and in particular to a method, apparatus, equipment, medium and product for self-calibration of power meters. Background Technology

[0002] With the rapid development of smart grids and distributed energy systems, higher demands are being placed on the accuracy, stability, and adaptability of power metering equipment in power systems. As core measurement terminals, power meters need to continuously output high-precision power parameters under complex power supply environments and possess the ability to perceive and dynamically adjust their own operating status in real time. Against this backdrop, edge computing technology has been gradually introduced into the field of power metering to assist in functions such as distributed data processing, anomaly detection, and system self-calibration. Related research largely focuses on the real-time analysis of voltage and current sampling signals by edge nodes, improving the fault tolerance of metering systems under scenarios such as power outages, phase drift, and time base jitter through local model building and data trend inference. Currently, some technologies explore power correction mechanisms based on edge models, but these are mostly limited to static coefficient adjustment and lack the ability to model the time-varying nature of electrical characteristics and channel heterogeneity, making it difficult to adapt to continuously changing power supply conditions.

[0003] Related technologies still face numerous limitations in remote status sensing and self-calibration of power meters. On the one hand, most solutions rely solely on calibration factors uniformly issued by the master station, lacking a local intelligent judgment mechanism at the meter level, thus failing to achieve accurate correction based on local physical conditions. On the other hand, most current calibration logics assume continuous and stable power supply, making it difficult to accurately reconstruct the offset path in scenarios of sudden power outages and power-on disturbances. Furthermore, existing clock correction strategies are relatively crude, failing to integrate historical offset trends and short-term fluctuation characteristics, significantly impacting the stability of the sampling frequency during long-term operation. Addressing these shortcomings, constructing an edge-side closed-loop adjustment mechanism with phase deviation sensing, power outage error compensation, and adaptive calibration capabilities has become a key technical challenge in the current intelligent upgrade process of power meters. Summary of the Invention

[0004] The purpose of this application is to provide a method, device, equipment, medium, and product for self-calibration of power meters, which can effectively sense phase deviation, perform power outage error compensation, and improve adaptive calibration capability.

[0005] To achieve the above objectives, this application provides the following solution:

[0006] Firstly, this application provides a self-calibration method for power meters, comprising:

[0007] Within the set sampling period of the edge node, the current power supply phase vector group is obtained based on the current power supply phase vector of each voltage channel in the collected power meter.

[0008] The current power supply phase vector group and the edge synchronization reference phase are processed by vector difference offset to obtain the comprehensive phase offset of the current cycle; the edge synchronization reference phase is an ideal phase obtained based on a short-cycle phase reference model preset inside the edge node.

[0009] After detecting a power supply interruption event, the zero-point error estimate is obtained based on the extreme value set of the power supply phase vector before the power outage and the initial power supply phase vector group after power is restored; the extreme value set of the power supply phase vector before the power outage is the extreme value point of each channel voltage in the cycle before the power outage.

[0010] The zero-point error estimate and the comprehensive phase offset are input into the calibration correction function to update the local correction factor group and form an adaptive calibration parameter group; the correction factor group includes the current transformer channel coefficient and the voltage conversion factor.

[0011] Secondly, this application provides a power meter self-calibration device, including: a data acquisition and processing module, a comprehensive phase offset module, a zero-point error estimation module, and a calibration module.

[0012] The acquisition and processing module is used to obtain the current power supply phase vector group based on the current power supply phase vector of each voltage channel in the acquired power meter within a set sampling period at the edge node.

[0013] The integrated phase offset module is used to perform vector difference offset processing on the current power supply phase vector group and the edge synchronization reference phase to obtain the integrated phase offset of the current period; the edge synchronization reference phase is an ideal phase obtained based on a short-period phase reference model preset inside the edge node.

[0014] The zero-point error estimation module is used to obtain the zero-point error estimation value based on the extreme value set of the power supply phase vector before the power outage and the initial power supply phase vector group after the power is restored after detecting a power outage abnormality interruption event; the extreme value set of the power supply phase vector before the power outage is the extreme value point of each channel voltage in the cycle before the power outage.

[0015] The calibration module is used to input the zero-point error estimate and the comprehensive phase offset into the calibration correction function, update the local correction factor group, and form an adaptive calibration parameter group; the correction factor group includes the current transformer channel coefficient and the voltage conversion factor.

[0016] Thirdly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the power meter self-calibration method described in any one of the above.

[0017] Fourthly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the power meter self-calibration method described in any one of the above descriptions.

[0018] Fifthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the power meter self-calibration method described above.

[0019] According to the specific embodiments provided in this application, the following technical effects are disclosed:

[0020] This application provides a self-calibration method, device, equipment, medium, and product for power meters. By acquiring the current power supply phase vector group and performing vector difference offset processing with the edge synchronous reference phase, the comprehensive phase offset of the current cycle can be accurately obtained, which facilitates improved accuracy. After detecting a power supply abnormality interruption event, the zero-point error estimation value can be calculated to effectively correct the phase deviation caused by power outage and power restoration, further improving the reliability of the system. The zero-point error estimation value and the comprehensive phase offset are jointly input into the calibration correction function to update the local correction factor group, forming a continuously evolving adaptive calibration parameter group with closed-loop adjustment capability. Through the adaptive calibration parameter group, the system can automatically complete the calibration process without relying on a central calibration source and manual intervention. It is suitable for remote deployment scenarios in complex environments and significantly improves the long-term operational stability and self-recovery capability of power meters. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is an application environment diagram of a power meter self-calibration method according to an embodiment of this application;

[0023] Figure 2 A flowchart illustrating a self-calibration method for power meters provided in an embodiment of this application;

[0024] Figure 3 This is a schematic diagram of the functional modules of a power meter self-calibration device provided in an embodiment of this application;

[0025] Figure 4 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application.

[0026] Figure label:

[0027] 1-Acquisition and processing module; 2-Comprehensive phase offset module; 3-Zero point error estimation module; 4-Calibration module. Detailed Implementation

[0028] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0029] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0030] The power meter self-calibration method provided in this application embodiment can be applied to, for example... Figure 1 In the application environment shown, terminal 101 communicates with server 102 via a network. A data storage system can store the data that server 102 needs to process. The data storage system can be set up independently, integrated into server 102, or placed in the cloud or on another server. Terminal 101 can send the current power supply phase vector group to server 102. Server 102 receives the current power supply phase vector group and performs vector difference offset processing on the current power supply phase vector group and the edge synchronization reference phase to obtain the comprehensive phase offset for the current period. The edge synchronization reference phase is an ideal phase obtained based on a short-cycle phase reference model preset within the edge node. After detecting a power supply interruption event, a zero-point error estimate is obtained based on the extreme value set of the power supply phase vector before the power outage and the initial power supply phase vector group after power restoration. The extreme value set of the power supply phase vector before the power outage consists of the extreme points of each channel voltage in the period before the power outage. The zero-point error estimate and the comprehensive phase offset are input into a calibration correction function to update the local correction factor group and form an adaptive calibration parameter group. The correction factor group includes the current transformer channel coefficient and the voltage conversion factor. Server 102 can feed back the obtained adaptive calibration parameter set to terminal 101. Furthermore, in some embodiments, the power meter self-calibration method can also be implemented by either server 102 or terminal 101 independently. For example, terminal 101 can directly perform power meter self-calibration on the current power supply phase vector set to be processed, or server 102 can obtain the current power supply phase vector set to be processed from the data storage system and perform power meter self-calibration on the current power supply phase vector set to be processed.

[0031] The terminal 101 can be, but is not limited to, various desktop computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, and smart in-vehicle devices. Portable wearable devices can include smartwatches, smart bracelets, and head-mounted devices. The server 102 can be implemented using a standalone server or a server cluster composed of multiple servers, or it can be a cloud server.

[0032] In one exemplary embodiment, such as Figure 2 As shown, a self-calibration method for power meters is provided. This method is executed by a computer device, specifically by a terminal or server alone, or by both a terminal and a server. In this embodiment, the method is applied to... Figure 1 Taking server 102 as an example, the explanation includes the following steps 201 to 204. Wherein:

[0033] Step 201: Within the set sampling period of the edge node, obtain the current power supply phase vector group based on the current power supply phase vector of each voltage channel in the collected power meter.

[0034] Step 202: Perform vector difference offset processing on the current power supply phase vector group and the edge synchronization reference phase to obtain the comprehensive phase offset of the current period; the edge synchronization reference phase is an ideal phase obtained based on a short-period phase reference model preset inside the edge node.

[0035] Step 203: After detecting a power supply interruption event, the zero-point error estimate is obtained based on the extreme value set of the power supply phase vector before the power outage and the initial power supply phase vector group after power-on; the extreme value set of the power supply phase vector before the power outage is the extreme value point of each channel voltage in the cycle before the power outage.

[0036] Step 204: Input the zero-point error estimate and the comprehensive phase offset into the calibration correction function to update the local correction factor group and form an adaptive calibration parameter group; the correction factor group includes the current transformer channel coefficient and the voltage conversion factor.

[0037] By implementing steps 201 to 204 above, the current power supply phase vector group is obtained, and vector difference offset processing is performed with the edge synchronous reference phase. This allows for accurate acquisition of the comprehensive phase offset of the current cycle, improving accuracy. After detecting a power supply interruption event, the zero-point error estimate is calculated to effectively correct the phase deviation caused by power outage and power restoration, further improving system reliability. The zero-point error estimate and the comprehensive phase offset are input into the calibration correction function to update the local correction factor group, forming a continuously evolving adaptive calibration parameter group with closed-loop adjustment capability. Through the adaptive calibration parameter group, the system can automatically complete the calibration process without relying on a central calibration source or manual intervention. This is suitable for remote deployment scenarios in complex environments, significantly improving the long-term operational stability and self-recovery capability of power meters.

[0038] In this application, the edge node periodically collects the real-time power supply phase vector of each voltage channel in the power meter. Specifically, within the set sampling period of the edge node, the multi-phase voltage channel measurement values ​​(i.e., the current power supply phase vector of each voltage channel) from the power meter are collected to construct the power supply phase vector group at the current moment (i.e., the current power supply phase vector group).

[0039] The current power supply phase vector group in step 201 of this application is not a simple vector set composed of instantaneous values ​​of voltage channels, but is obtained by performing phase calculation processing on each voltage signal. Specifically, this includes normalization, period determination, zero-crossing point extraction, and main phase extraction.

[0040] Formula expression: Given a three-phase voltage signal u A (t), u B (t), u C (t), then the principal phase value of each phase can be expressed as:

[0041] in, For voltage signal u i (t) is subjected to Hilbert transform to obtain the analytic signal; arg(·) is the extracted instantaneous phase angle (dominant phase); t k This represents the current sampling time.

[0042] Therefore, the current power supply phase vector set is: Θ(t) k )=[φ A (t k ), φ B (t k ), φ C (t k )] T .

[0043] In this process, the edge nodes use a locally configured virtual PLL (Phase-Locked Loop) algorithm or phase calculation model to perform amplitude normalization and period determination. Based on the changing trend of zero-crossing points between sampling points, they solve for the principal phase value of each channel in the current period. This principal phase value is the instantaneous phase angle extracted from the current real-time sampled voltage signal, representing the voltage phase measured by the physical quantity. This phase value can be represented by the envelope principal angle obtained by the sine fitting angle of the voltage channel or the Hilbert transform.

[0044] Because three-phase power systems are subject to external influences such as load imbalance, cable impedance asymmetry, and voltage disturbances, using only single-channel phase data to determine the presence of sampling drift can easily lead to misjudgments. Therefore, by constructing the above structure, the phase state can be expressed with multi-channel consistency, which helps to identify systematic distortions in the sampling reference.

[0045] In another exemplary embodiment of this application, in order to improve accuracy and effectively sense phase deviation, the above step 202 is replaced by the following steps 301 to 302:

[0046] Step 301: Based on the short-cycle phase reference model pre-set inside the edge node, calculate the component vector angle difference of the current power supply phase vector group to obtain the difference vector group; each component of the difference vector group is the phase deviation angle.

[0047] Step 302: Average the phase deviation angle to obtain the comprehensive phase offset for the current cycle.

[0048] In step 301, a short-cycle phase reference model is pre-set inside the edge node (i.e., the short-cycle phase reference model fixed in the edge node), which is used to represent the voltage phase structure that should be presented under the local load topology under ideal power supply conditions. It is constructed from the sampled values ​​under static load within the power frequency cycle and maintains stability for one unit time. During the actual sampling phase, the current power supply phase vector set and the short-cycle phase reference model are analyzed component by component, i.e., the difference is calculated, to construct the difference vector set.

[0049] Specifically, the current power supply phase vector set is: Θ(t k )=[φ A (t k ), φ B (t k ), φ C (t k )] T .

[0050] The short-period phase reference model is as follows:

[0051] The difference vector set formed by calculating the angle difference between the component vectors is:

[0052]

[0053] Where, Δφ i The phase offset angle for each phase, The phase offset angle can be normalized to the range of (-π, π), which facilitates periodic processing.

[0054] This set of difference vectors represents the degree of deviation of each channel voltage from the ideal phase value at the current moment. By calculating the difference in the angle between component vectors, the periodic phase shift caused by instrument clock drift can be located. This method avoids frequency synchronization failures caused by traditional PLL unlocking or external synchronization module errors, and can independently complete high-precision modeling of phase errors under power supply conditions at edge nodes.

[0055] In step 302, the phase deviation angles (i.e., the three-phase offset values) are averaged to construct the comprehensive phase offset for the current cycle. Specifically, the phase deviation angle is Δφ. A , Δφ B , Δφ C Then the overall phase offset (mean) is:

[0056]

[0057] Where, Δφ avg (t k ) is the comprehensive phase offset for the current cycle, used to reflect the overall systematic phase drift of the instrument sampling system within this cycle.

[0058] The comprehensive phase offset is calculated by performing three-phase balance offset analysis (i.e., averaging) on ​​the phase deviation angle (or phase deviation angle value); this comprehensive phase offset is used to determine whether there is a systematic phase shift characteristic in the current instrument's local sampling reference.

[0059] In conventional designs, many instrument manufacturers use a single-phase reference as the sampling clock criterion, only making reference determinations. This can easily mislead the controller into making unnecessary sampling corrections when there are phase-to-phase imbalances or single-phase fluctuations in the power grid. This invention sets a multi-phase average as the criterion, improving the robustness of identifying systematic sampling drift, enabling the instrument to tolerate short-period phase fluctuations, and enhancing stable calibration response capabilities in multi-channel mixed distortion environments.

[0060] In another exemplary embodiment of this application, in order to effectively respond to abnormal power supply interruption events and perform power outage error compensation, the above step 203 is replaced by the following steps 401 to 406:

[0061] Step 401: After detecting a power supply interruption event, record the power supply phase vector sequence in the cycle before the power outage.

[0062] Step 402: Based on the extreme points of each channel voltage extracted from the power supply phase vector sequence, obtain the extreme value set of the power supply phase vector before power failure.

[0063] Step 403: After the power-on recovery event is triggered, a stable static segment is determined based on the starting segment in the sampling reference sequence after power-on.

[0064] Step 404: Calculate the initial average phase value set in the stable static segment from the initial power supply phase vector set extracted after power-on.

[0065] Step 405: Perform channel-level matching difference between the extreme value set of the power supply phase vector before power failure and the initial average phase value set to obtain the three-phase voltage zero-point offset.

[0066] Step 406: Perform sliding window smoothing on the three-phase voltage zero-point offset to obtain the zero-point error estimate for the current cycle.

[0067] After detecting a power outage, the edge node records the extreme value set of the power supply phase vector before the power outage in the previous cycle. Then, in the static segment after the power restoration event is triggered, the initial power supply phase vector group after power restoration is re-extracted. The extreme value set of the power supply phase vector before the power outage in the previous cycle and the re-extracted initial power supply phase vector group after power restoration are compared to establish a zero-point error estimation.

[0068] In steps 401 to 402, when the edge node detects a power supply interruption event, it records the power supply phase vector sequence (i.e., the power supply phase vector sequence in the cycle before the power outage) within multiple consecutive time points before the event occurs, and extracts the extreme points of each channel voltage to form the extreme value set of the power supply phase vector before the power outage.

[0069] Specifically, in the edge sampling control module, a power supply status monitoring threshold is set within a continuous sampling period to identify power grid interruption events. The edge node collects the current power supply phase vector group Θ(t) in real time within the periodic time interval. kIn this three-phase voltage vector group, each element represents the principal phase value of a channel. The system uses continuous window differences to determine if there are situations such as a simultaneous sudden drop in voltage amplitude across all channels, zero-crossing failure, or cessation of phase fluctuations. If the absolute value of the continuous window differences across multiple consecutive sampling points is less than the average threshold, it is marked as a power outage event. The window data prior to this outage event is used to construct a set of phase extrema before the power outage. The system also searches for the maximum value point of each channel to construct an extrema set. This extrema set is stored in an event buffer as a reference for the voltage state before the power outage. Unlike conventional interruption processing that only records amplitude or interruption time, this application emphasizes "preservation of waveform structure continuity," ensuring the completeness of the electrical parameter hierarchy in the state before the power outage.

[0070] Constructing the extreme value set of the power supply phase vector before power failure in steps 401-402 is one of the key strategies for achieving "waveform structure continuity preservation" in this application. Although the extreme points appear discrete, they can fully preserve the temporal characteristics and structural trends of the original waveform in electrical parameter modeling. In periodic AC signals, the maxima and minima reflect the symmetrical structure and phase characteristics of each cycle. The temporal intervals, amplitude comparisons, and order of occurrence between them can basically restore the "shape trend" of the waveform. If multiple extreme points are sampled within a continuous cycle (such as a multi-cycle extreme value set), the waveform phase trend and main frequency stability of each channel can be deduced, which is more structurally expressive than single-point amplitude or frequency recording. Even if the extreme points are discrete, by preserving the relative positions and phase angle information of key points, it is possible to track whether the waveform shape has drifted or become distorted.

[0071] In steps 403 to 404, after the power-on recovery event is triggered, a stable static segment is constructed in the starting segment of the sampling reference sequence, the power supply phase vector sequence with the same structure is extracted, and the initial average phase value group of the segment is obtained.

[0072] Specifically, when the edge sampling control module (or edge controller) detects power restoration, i.e., when the three-phase voltages once again meet the normal operating amplitude and phase fluctuation conditions, it begins to construct the initial time window segment for power restoration, denoted as [t]. r , t r+m ], where t r This indicates the start time point for recovery sampling, and variable m is the length of the stable window. Within this time window, the initial power supply phase vector group (or sampled phase vector sequence) after power-on is extracted:

[0073] Θ(t r +j)={θ a (t r +j), θ b (t r +j), θ c (t r+j)}, j=0,1,...,m.

[0074] Wherein, Θ(t) r +j) indicates at time point t r The three-phase voltage phase vector set under +j is a three-dimensional phasor set used to describe the phase state of the three-phase voltage at that moment; θ a (t r +j) indicates at time point t r The phase angle value of phase a in the three-phase voltage under +j condition; θ b (t r +j) indicates at time point t r The phase angle value of phase b in the three-phase voltage under +j condition, θ c (t r +j) indicates at time point t r +j represents the phase angle value of phase c in the three-phase voltage, where j indicates the time window [t]. r , t r The sampling point index variable within [+m] takes values ​​from 0 to m.

[0075] The weighted average is used to extract the average static phase of the three phases, and an initial average phase value set is constructed (or the initial state reference vector is restored).

[0076] Compared to the single-point sampling initialization method after power restoration in conventional power systems, the current design constructs a multi-point mean model in the early stage of restoration, which effectively suppresses the instantaneous disturbance of power supply connection and ensures that the subsequent zero-point offset comparison benchmark has time stability and fluctuation buffering capability.

[0077] In step 405, channel-level matching difference analysis is performed on the extreme value set of the power supply phase vector before power failure and the initial average phase value set to construct the three-phase voltage zero-point offset.

[0078] Specifically, using the extreme value set of the power supply phase vector before the power outage and the initial average phase value set as inputs, the difference is calculated for each channel to obtain the three-phase voltage zero-point offset (vector).

[0079] This difference reflects the relative displacement between the voltage vector extremum point (or the set of extrema of the power supply phase vector before power failure) and the newly established initial zero point (or the initial average phase value set). This process does not depend on voltage peak amplitude or frequency variations, but only on the position of the main phase angle in the waveform structure, avoiding erroneous offset judgments caused by amplitude disturbances, load inrush currents, or power factor changes.

[0080] In step 406, the three-phase voltage zero-point offset is smoothed by a sliding window to obtain the estimated value of the (stable) zero-point error in the current cycle, and it is cached in the zero-point offset buffer as the basis for subsequent coefficient updates.

[0081] Because there may be brief fluctuations or residual interference in the initial stage after power-on, there is a stability risk when directly using the three-phase voltage zero-point offset vector ΔΘ0 as the calibration basis. Therefore, a sliding window smoothing mechanism is constructed. The historical zero-point error estimation queue is as follows:

[0082]

[0083] Where L represents the sliding window width, typically 3 to 5 recovery periods. The weighted moving average model is used to extract the current stable zero-point error estimate (or stable zero-point offset):

[0084]

[0085] Among them, w k This represents the weighting coefficient for the corresponding period. It can generally be set to a linearly decreasing or exponentially decreasing form, prioritizing the contribution of the latest period data and improving compatibility with volatility suppression and trend response.

[0086] The final calculation The data is written to a dedicated zero-point offset buffer. The data in the buffer will serve as direct input for the transformation factor update in the next step, S4. By employing a sliding smoothing model, stable and continuously trackable three-phase zero-point offset values ​​are constructed. Compared to existing single-cycle comparison strategies, this method exhibits greater time robustness and multi-cycle compatibility, enhancing the self-recovery capability of edge nodes after abnormal events.

[0087] To enhance the intelligent collaboration capabilities of edge nodes, this invention further encapsulates the phase offset data, calibration coefficients, and abnormal event information (such as power outage flags and sudden shifts in offset trends) extracted during edge self-calibration into a status description data packet. This packet is then uploaded to the upper-level platform or power master station system via standardized communication protocols (such as MQTT, IEC 104, or Modbus TCP), enabling remote status awareness of terminal instruments. The status description data packet can serve as one of the reference indicators for assessing the health of the power grid edge in the master station platform, and can be used for subsequent centralized management, early warning response, and maintenance scheduling.

[0088] In another exemplary embodiment of this application, in order to effectively improve the adaptive calibration capability, the zero-point error estimate and the comprehensive phase offset are jointly input into the calibration correction function to update the local correction factor group and form an adaptive calibration parameter group for subsequent cycle power measurement updates. The above step 204 is replaced by the following steps 501 to 506:

[0089] Step 501: Construct a calibration input vector based on the zero-point error estimate and the comprehensive phase offset of the current cycle.

[0090] Step 502: Input the calibration input vector into the calibration correction function built locally at the edge node, and output the correction factor group for the current cycle for each phase channel.

[0091] Step 503: Compare the correction factor group stored in the previous period with the correction factor group in the current period.

[0092] Step 504: If the change amplitude is lower than the set parameter disturbance threshold, then update to the correction factor group for the current period.

[0093] Step 505: If the change amplitude is not lower than the set parameter disturbance threshold, then record it in the candidate factor queue.

[0094] Step 506: The updated set of correction factors is used as the input benchmark for the next cycle, and an adaptive calibration parameter set is constructed.

[0095] In step 501, the edge node retrieves the estimated zero-point error value within the current period (stable). and combined phase offset Δφ avg (t k ), construct the calibration input vector, which serves as the input parameter set for the calibration correction function.

[0096] Specifically, the input structure of each phase channel x∈{a, b, c} is defined as: (where (a\b\c) are channels).

[0097]

[0098] Among them, E x For each phase channel input vector, This is the estimated value of the zero-point error, Δφ avg (t k ) represents the combined phase offset.

[0099] The calibration input vector set {E a E b E c The input group is organized as the overall correction function. In existing technologies, electricity metering devices mostly rely on statically set scaling factors and cannot dynamically respond to changes in edge physical states. The current design models structural phase shifts and time-domain errors together, possessing cross-domain calibration capabilities, and is an innovative extension of traditional calibration coefficient adjustment schemes.

[0100] In step 502, the calibration input vector is input to the calibration correction function built locally at the edge node, and the correction factor group for the current cycle, including the current transformer channel coefficient and voltage conversion factor, is output for each phase channel.

[0101] The input vector E for each phase channel mentioned above xThe input to the local embedded (channel) calibration correction function is defined as follows:

[0102]

[0103] in, Indicates the current period t i Calibration coefficient of current transformer channel x; This represents the voltage conversion scaling factor for the corresponding channel. The function is internally constructed based on a two-factor nonlinear perturbation matching structure, for example:

[0104]

[0105] Among them, γ1, γ2, λ1, and λ2 are adjustment coefficients used to control the error mapping ratio, and each parameter is set with consideration of signal amplification suppression conditions to avoid sudden changes in coefficients.

[0106] This approach enables the electrical parameter factor to evolve periodically and is non-statically loaded, significantly improving the accuracy of power measurement and resistance to long-term drift under marginal instability conditions.

[0107] In steps 503 to 505, the correction factor set stored in the previous period is compared with the correction factor set in the current period. If the change is less than the set parameter disturbance threshold, the correction factor set in the current period is updated to the local measurement module.

[0108] To avoid coefficient oscillations caused by short-term disturbances or false triggering events, a periodic disturbance verification mechanism is set up to compare the change magnitude of the stored parameters (or the correction factor set stored in the previous period) with the current period's output parameters (or the correction factor set of the current period). The change difference is defined as:

[0109]

[0110] The correction factor set update for the current period is allowed to be written to the local measurement module when the following conditions are met:

[0111]

[0112] Where, ε I and ε V This indicates the preset disturbance limit threshold, typically fluctuating between 0.1% and 0.5%. This mechanism prevents short-term oscillations from propagating to downstream measurement models and causing cumulative electrical errors; it is a dynamically adjustable stabilizer. Compared to traditional calibration methods that employ periodic manual verification or constant loading, this structure possesses "self-sensing of change" and "tolerance for gradual changes" characteristics.

[0113] If the current factor difference exceeds the above threshold, the edge node will not update the measurement model for the time being, but will only record the current calculation result (or the correction factor group of the current period) into the candidate factor queue, and continue to perform dynamic comparison in the next period.

[0114] In another exemplary embodiment of this application, in order to effectively improve the adaptive calibration capability, an offset regression model between voltage and current sampling clocks is established at the edge side based on the comprehensive phase offset, and the local sampling time series is updated to construct a corrected time reference series, specifically including the following steps 205 to 208. Wherein:

[0115] Step 205: Based on the comprehensive phase offset, establish an offset regression model between the voltage and current sampling clocks on the edge side.

[0116] Step 206: Based on the offset regression model, predict the expected sampling offset for the current period.

[0117] Step 207: Based on the original sampling time series of the current period and the expected sampling offset of the current period, obtain the corrected time series.

[0118] Step 208: Use the correction factor set and the corrected time series as the input reference for the next period to jointly construct the adaptive calibration parameter set.

[0119] According to steps 205 to 208, this application updates the local sampling time series based on the original sampling time series of the current period and the expected sampling offset of the current period, and replaces the original sampling points with the corrected time base as a reference for future sampling and calibration, thereby improving calibration accuracy and enhancing the accuracy and reliability of the adaptive calibration mechanism.

[0120] In another exemplary embodiment of this application, in order to facilitate the correction of the sampling time series (i.e., time reference) on the edge side, the above step 205 is replaced by the following steps 601 to 603:

[0121] Step 601: Add the current cycle's composite phase offset to the historical sampling offset queue.

[0122] Step 602: Extract the phase offset values ​​within the sliding window from the historical sampling offset queue, and construct the observation vector and its corresponding time label.

[0123] Step 603: Based on the observation vector and time label, construct an offset regression model between the voltage and current sampling clocks.

[0124] In step 601, the comprehensive phase offset in the current period is added to the historical sampling offset queue, and the offset trend function in the sliding window is calculated to provide a trend input basis for subsequent self-calibration.

[0125] Specifically, the overall phase offset Δφ avg (t k The value representing the overall sampling offset within a single period is added to the historical sampling offset queue (or historical sampling offset buffer queue), which is defined as follows: Here, w represents the width of the sliding window, typically set to 3–5 sampling periods, used to dynamically model the sampling offset trend over a recent period. Compared to existing methods that use the mean or median as representative quantities for static adjustment, constructing a time-sliding offset queue can capture the changing trend of sampling error over multiple periods.

[0126] To further describe this trend change, a one-dimensional offset trend function within the sliding window is constructed using simple fitting or weighted moving difference methods. The historical offset sequence within the sliding window is defined as follows:

[0127]

[0128] Where n is the window width.

[0129] A simple linear fit (least square regression) can be used to construct the regression function:

[0130] △φ avg (t) = at + b.

[0131] Where t is the sample time number; a is the slope of the offset trend, representing the rate of change of offset within a unit period; and b is the intercept, representing the initial offset baseline. The fitting method is least squares fitting, i.e.:

[0132]

[0133] Among them, y i =Δφ avg (t i ), which is a regression function, and the regression coefficients indicate the rate of change of the current sampling offset and the initial offset reference, respectively.

[0134] In the regression process, introducing weighting factors can further enhance the ability to suppress short-term outliers. For example, setting the weight of recent time period values ​​higher than that of distant time period values ​​can control the model's responsiveness to mutations.

[0135] In step 602, the phase offset values ​​within the sliding window are extracted from the historical sampling offset queue, and the observation vector and its corresponding time label are constructed as input data for regression modeling.

[0136] Specifically, the conditions for extracting phase offset values ​​within the sliding window include the following aspects: A fixed-length sliding window is typically set, and the comprehensive phase offset corresponding to the most recent multiple periods is selected; if phase anomalies, signal loss, or synchronization loss are detected in certain periods, the corresponding offset should be skipped. Alternatively, median filtering, IQR (Interquartile Range) methods, or Z-scores can be used to judge and remove outliers; if a large phase jump occurs in the queue, it indicates a sudden change in the system state, and the window may not be representative, so prediction can be temporarily frozen; the sampling rate and sampling time corresponding to the selected period should be consistent, otherwise it cannot be used for trend regression.

[0137] Since step 202 yields the comprehensive phase offset for the current period, reflecting the offset phenomenon at a specific moment, system calibration and correction require predicting whether the offset will continue at the "next moment," i.e., identifying the "offset trend." Therefore, a time series trend model must be established using offsets from multiple historical periods to achieve "prediction" and "regression modeling." Furthermore, using only the current value may be misled by abnormal interference such as short-term noise, power grid disturbances, and instrument glitches, leading to over-correction or repeated oscillations.

[0138] In step 603, a regression function is constructed based on the observation vector and time label to predict the expected sampling offset of the next sampling period. The regression function uses a multi-order weighted difference function to express the sampling drift trend.

[0139] To identify the current phase shift trend and predict future period sampling errors, the regression model function is defined as follows:

[0140]

[0141] Among them, f τ (t) represents the expected sampling time offset at the current time point t; variable Δ k δ θ (t) represents the phase shift change in the k-th order difference form, i.e.:

[0142] Δ 1 δ θ (t k )=δ θ (t k )-δ θ (t k-1 ).

[0143] Δ 2 δ θ (t k )=Δ 1 δ θ (t k )-Δ1 δ θ (t k-1 ).

[0144] And so on up to order k. Variable α k The weighting coefficients for the k-th order difference component are set based on empirical values, historical fluctuation amplitudes, and sampling system inertia. Compared with conventional linear regression models, this method retains the multi-order dynamic characteristics of the shift changes in the model structure, and can still maintain stable prediction capabilities even when the phase drift exhibits high-order nonlinear changes.

[0145] The use of multi-order difference models can also be used to reduce the impact of low-frequency noise, enabling the system to maintain its predictive ability even when there are long-period power frequency interferences or low-amplitude voltage disturbances.

[0146] In steps 206-207, the original sampling time series of the current period and the predicted expected sampling offset of the current period are combined to construct the corrected time series (or sampling time), and the sampling time series is updated in the local edge sampling controller. This time series covers the pre-calibrated sampling points of the next n periods.

[0147] Specifically, the current period t i+1 The corresponding expected sampling offset estimate is: δτ i+1 =f τ (t i+1 ).

[0148] Current period t i+1 The corresponding original sampling time series is set as: T = {t} i+1 , t i+2 , ..., t i+n};

[0149] Here, variable n represents the number of subsequent predictive control cycles. This is combined with the expected sampling offset estimate δτ obtained from the previous prediction step. i+1 Calculate the corrected sampling time point: t′ j =t j +δτ j (j∈[i+1, i+n]),

[0150] The above formula shows that the subsequent n sampling periods all introduce dynamic offsets based on the original planned time, forming the corrected time series: T′={t′ i+1 ,t′ i+2 , ..., t′ i+n}

[0151] The sampling controller uses this sequence as a time reference for triggering A / D sampling and vector construction in future cycles. Unlike existing periodic sampling control based on the timer interrupt mode of the main control MCU (Microcontroller Unit), the current method supports soft calibration and dynamic sampling point replacement, adapts to the current cycle sensing results for offset compensation, and significantly reduces the risk of off-sampling caused by system crystal oscillator errors, phase-locked loop mistracking, and external synchronization errors.

[0152] By adopting an adaptive update mechanism for sampling time, the sampling strategy can be autonomously adjusted without relying on traditional GPS synchronization systems or master station calibration commands, providing time accuracy support for the next stage of measurement parameter correction.

[0153] In addition, after obtaining the corrected time series, the original sampling time series and the corrected time series are compared, the sampling frequency adjustment vector is calculated, and it is stored in the sampling frequency dynamic adjustment buffer for error closed-loop verification and long-term sampling drift trend analysis in subsequent periods.

[0154] Specifically, the corrected time series T′ is subjected to element-wise difference analysis with the original time series T to construct the frequency adjustment vector for the current period:

[0155]

[0156] Among them, variable T s This represents the preset theoretical sampling interval period, in seconds. Each term Δf in the adjustment vector... j This represents the frequency offset rate of the current sampling point, which is written as a frequency fine-tuning factor in the name... Dynamically adjust the cache:

[0157]

[0158] This buffer provides two functions: First, it provides a base frequency correction factor for subsequent cycles to maintain the stability of the calibration strategy when there are no on / off disturbances in the system; Second, it serves as a historical sampling offset feedback source to construct long-term drift trend statistical indicators, such as the standard deviation change rate and the maximum continuous offset window, which are used to dynamically adjust the regression model window width, difference order, and weight coefficients.

[0159] This invention overcomes the limitations of existing sampling systems that rely on master station commands, linear models, and offline calibration. It achieves autonomous benchmark updates within distributed edge computing nodes, demonstrating strong novelty and practical engineering feasibility. Further development of parameter adjustment strategies, anomaly calibration shielding mechanisms, or periodic stability assessments can be achieved by extending the control logic based on the current output correction sequence and frequency vector.

[0160] In step 208, the updated correction factor set and the corrected time series (or the current corrected time series) are loaded together into the instrument measurement control logic as the input reference for the next cycle's power measurement and edge sensing module, thus completing the self-calibration closed-loop update for this cycle.

[0161] This update strategy differs from conventional static calibration, featuring a complete closed-loop process of "measurement-error identification-parameter update-measurement," and all update processes do not rely on master station intervention or external standard voltage source input.

[0162] Furthermore, in the edge node environment, a parameter evolution chain of "input data change - controllable model output - adjustable update rhythm" is added to solve the problems of difficult self-calibration, slow parameter drift repair, and weak adaptability in existing power meters. The system can operate for a long time in a distributed edge environment that cannot be connected to a central calibration system, achieving continuous self-sensing and self-adjusting power measurement capabilities.

[0163] This invention, based on an edge analysis architecture, addresses key issues in power meters such as sampling timing errors, power supply phase offsets, and zero-point drift caused by on / off anomalies. It proposes a remote state awareness and self-calibration method that integrates phase difference analysis, sampling clock regression modeling, and channel-level adaptive correction. By periodically extracting the power supply phase vector at edge nodes and analyzing its angle difference with the synchronization reference, a sampling deviation trend is constructed. Furthermore, a regression model based on historical offsets is established to dynamically adjust the local sampling time series, significantly improving time base accuracy. In power-on anomaly events, this method can extract the phase extreme value structure before and after power outages, constructing zero-point error estimates and effectively compensating for unplanned offsets. Simultaneously, multi-source deviations are used as inputs to drive the local correction function, updating the current transformer coefficients and voltage conversion factors to form a continuously evolving calibration parameter set with closed-loop adjustment capabilities. This method does not rely on a central calibration source or manual intervention, making it suitable for remote deployment scenarios in complex environments and significantly improving the long-term operational stability and self-recovery capability of power meters.

[0164] Based on the same inventive concept, this application also provides a power meter self-calibration device for implementing the power meter self-calibration method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations of one or more power meter self-calibration device embodiments provided below can be found in the limitations of the power meter self-calibration method described above, and will not be repeated here.

[0165] In one exemplary embodiment, such as Figure 3 As shown, a power meter self-calibration device is provided, including: a data acquisition and processing module 1, a comprehensive phase offset module 2, a zero-point error estimation module 3, and a calibration module 4.

[0166] The acquisition and processing module 1 is used to obtain the current power supply phase vector group based on the current power supply phase vector of each voltage channel in the acquired power meter within a set sampling period of the edge node.

[0167] The integrated phase offset module 2 is used to perform vector difference offset processing on the current power supply phase vector group and the edge synchronization reference phase to obtain the integrated phase offset of the current period; the edge synchronization reference phase is an ideal phase obtained based on the short-period phase reference model preset inside the edge node.

[0168] The zero-point error estimation module 3 is used to obtain the zero-point error estimation value based on the extreme value set of the power supply phase vector before the power outage and the initial power supply phase vector group after the power is turned on after detecting a power supply abnormal interruption event; the extreme value set of the power supply phase vector before the power outage is the extreme value point of each channel voltage in the cycle before the power outage.

[0169] The calibration module 4 is used to input the zero-point error estimate and the comprehensive phase offset into the calibration correction function, update the local correction factor group, and form an adaptive calibration parameter group; the correction factor group includes the current transformer channel coefficient and the voltage conversion factor.

[0170] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 4 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores power meter self-calibration data. The I / O interfaces are used for information exchange between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a power meter self-calibration method.

[0171] Those skilled in the art will understand that Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0172] In one exemplary embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0173] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0174] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0175] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0176] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0177] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0178] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0179] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A self-calibration method for power meters, characterized in that, include: Within the set sampling period of the edge node, the current power supply phase vector group is obtained based on the current power supply phase vector of each voltage channel in the collected power meter. The vector difference offset is processed by the current power supply phase vector group and the edge synchronization reference phase to obtain the comprehensive phase offset of the current cycle; The edge synchronization reference phase is an ideal phase obtained based on a short-period phase reference model preset inside the edge node; After detecting a power supply interruption event, the zero-point error estimate is obtained based on the extreme value set of the power supply phase vector before the power outage and the initial power supply phase vector group after power is restored; the extreme value set of the power supply phase vector before the power outage is the extreme value point of each channel voltage in the cycle before the power outage. The zero-point error estimate and the comprehensive phase offset are input into the calibration correction function to update the local correction factor group and form an adaptive calibration parameter group; the correction factor group includes the current transformer channel coefficient and the voltage conversion factor.

2. The self-calibration method for power meters according to claim 1, characterized in that, The vector difference offset is processed from the current power supply phase vector group and the edge synchronization reference phase to obtain the comprehensive phase offset for the current cycle, specifically including: Based on the short-cycle phase reference model pre-set inside the edge node, the component vector angle difference of the current power supply phase vector group is calculated to obtain the difference vector group; each component of the difference vector group is the phase deviation angle. The phase deviation angle is averaged to obtain the comprehensive phase offset for the current cycle.

3. The self-calibration method for power meters according to claim 1, characterized in that, After detecting a power outage, the zero-point error estimate is obtained based on the extreme value set of the power supply phase vector before the power outage and the initial power supply phase vector set after power restoration. Specifically, this includes: After detecting a power supply interruption event, record the power supply phase vector sequence in the cycle before the power outage; Based on the extreme points of each channel voltage extracted from the power supply phase vector sequence, the extreme value set of the power supply phase vector before the power outage is obtained; After the power-on recovery event is triggered, a stable static segment is determined based on the starting segment of the post-power-on sampling reference sequence; The initial power supply phase vector set extracted from the stable static segment is used to calculate the initial average phase value set of the stable static segment; The channel-level matching difference between the extreme value set of the power supply phase vector before the power outage and the initial average phase value set is used to obtain the zero-point offset of the three-phase voltage. The zero-point offset of the three-phase voltage is smoothed by a sliding window to obtain the zero-point error estimate for the current cycle.

4. The self-calibration method for power meters according to claim 1, characterized in that, The zero-point error estimate and the combined phase offset are input into the calibration correction function to update the local correction factor set and form an adaptive calibration parameter set, which specifically includes: Based on the zero-point error estimate and the overall phase offset of the current cycle, a calibration input vector is constructed; The calibration input vector is input to the calibration correction function built locally at the edge node, and the correction factor group for the current cycle is output separately for each phase channel. Compare the correction factor set stored in the previous period with the correction factor set in the current period; If the change is less than the set parameter disturbance threshold, then update to the correction factor group for the current period; If the change is not less than the set parameter disturbance threshold, it is recorded in the candidate factor queue; The updated set of correction factors is used as the input benchmark for the next cycle, and an adaptive set of calibration parameters is constructed.

5. The self-calibration method for power meters according to claim 1, characterized in that, Also includes: Based on the comprehensive phase offset, an offset regression model between voltage and current sampling clocks is established at the edge side; Based on the offset regression model, predict the expected sampling offset for the current period; Based on the original sampling time series of the current period and the expected sampling offset of the current period, the corrected time series is obtained; The set of correction factors and the corrected time series are used as the input reference for the next period to jointly construct the adaptive calibration parameter set.

6. The power meter self-calibration method according to claim 5, characterized in that, Based on the comprehensive phase offset, an offset regression model between the voltage and current sampling clocks is established at the edge side, specifically including: Add the current cycle's composite phase offset to the historical sampling offset queue; Extract the phase offset values ​​within the sliding window from the historical sampling offset queue to construct the observation vector and its corresponding time label; Based on the observation vector and time label, an offset regression model between voltage and current sampling clocks is constructed.

7. A self-calibration device for power meters, characterized in that, include: The data acquisition and processing module is used to obtain the current power supply phase vector group based on the current power supply phase vector of each voltage channel in the acquired power meter within a set sampling period at the edge node. The integrated phase offset module is used to perform vector difference offset processing on the current power supply phase vector group and the edge synchronization reference phase to obtain the integrated phase offset of the current period; the edge synchronization reference phase is an ideal phase obtained based on a short-period phase reference model preset inside the edge node; The zero-point error estimation module is used to obtain the zero-point error estimation value based on the extreme value set of the power supply phase vector before the power outage and the initial power supply phase vector group after the power is restored after detecting a power outage abnormality interruption event; the extreme value set of the power supply phase vector before the power outage is the extreme value point of each channel voltage in the cycle before the power outage; The calibration module is used to input the zero-point error estimate and the comprehensive phase offset into the calibration correction function, update the local correction factor group, and form an adaptive calibration parameter group; the correction factor group includes the current transformer channel coefficient and the voltage conversion factor.

8. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the power meter self-calibration method according to any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the power meter self-calibration method according to any one of claims 1-6.

10. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the power meter self-calibration method according to any one of claims 1-6.

Images