Adaptive calibration method, device and equipment of metering box and storage medium

Abstract

This invention relates to the field of metrology calibration, and more particularly to an adaptive calibration method, apparatus, device, and storage medium for a metrology box. The method includes the following steps: powering on the metrology box normally, performing self-operating noise detection, and constructing a self-operating noise feature set; performing initial calibration and real-time metrological detection on the metrology box based on the self-operating noise feature set, outputting multiple metrological data sequences; performing sequence-by-sequence feature analysis and differential calculation of adjacent time periods based on the multiple metrological data sequences to obtain a feature difference value sequence; synchronously acquiring environmental monitoring parameters of the metrology box; tracing environmental interference sources based on the environmental monitoring parameters and marking interference sources; performing offset calibration based on the interference sources to obtain offset calibration reference parameters; and performing real-time operating condition metrological detection based on the offset calibration reference parameters. This invention achieves accurate and efficient metrology box detection and calibration.

Description

Technical Field

[0001] This invention relates to the field of metrology and calibration, and in particular to an adaptive calibration method, apparatus, device, and storage medium for a metrology box. Background Technology

[0002] During long-term operation, metering boxes are susceptible to parameter drift and measurement errors due to various factors such as load fluctuations, changes in ambient temperature and humidity, electromagnetic interference, and component aging, leading to a gradual decrease in measurement accuracy. Furthermore, in complex power environments, sudden load changes or abnormal operating conditions can cause instantaneous anomalies in measurement data, further exacerbating the instability of measurement results. If these problems are not detected and corrected in a timely manner, they may result in inaccurate electricity metering, potentially leading to economic disputes or even affecting the normal operation of the power system. Existing metering box calibration methods mainly rely on periodic manual verification or offline calibration, typically calibrating parameters within fixed cycles. However, after the metering box is put into operation, its operating environment and load conditions often exhibit continuous changes. Traditional static calibration methods struggle to respond to these changes promptly, easily resulting in significant measurement deviations between calibration cycles. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention proposes an adaptive calibration method, apparatus, device, and storage medium for a metrology box, thereby resolving at least one of the aforementioned technical problems.

[0004] To achieve the above objectives, the present invention provides an adaptive calibration method for a metrology box, comprising the following steps: Step S1: Power on the metering box normally, perform self-operation noise detection, and construct a self-operation noise feature set; Step S2: Perform initial calibration and real-time measurement detection on the metering box according to the self-running noise feature set, and output multiple measurement data sequences; Step S3: Perform sequence-by-sequence feature analysis and difference calculation between adjacent time periods based on the multiple measurement data sequences to obtain a sequence of feature difference values; Step S4: Synchronously collect environmental monitoring parameters of the metering box; trace the source of environmental interference based on the characteristic difference value sequence according to the environmental monitoring parameters, and mark the interference source; Step S5: Perform offset calibration based on the interference source to obtain offset calibration reference parameters; perform real-time operating condition measurement and detection based on the offset calibration reference parameters.

[0005] This specification provides an adaptive calibration device for a metrology box, used to perform the adaptive calibration method for a metrology box as described above, comprising: The noise detection unit is used to drive the metering box to power on normally, perform self-operation noise detection, and construct a self-operation noise feature set. The metering unit is used to perform initial calibration and real-time metering detection on the metering box according to the self-operating noise feature set, and output multiple metering data sequences. The difference calculation unit is used to perform sequence feature analysis and difference calculation between adjacent time periods based on the multiple measurement data sequences to obtain a sequence of feature difference values. The interference tracing unit is used to synchronously collect environmental monitoring parameters from the metering box; and to trace the source of environmental interference based on the characteristic difference value sequence according to the environmental monitoring parameters, and mark the interference source. The calibration unit is used to perform offset calibration based on the interference source to obtain offset calibration reference parameters; and to perform real-time operating condition measurement and detection based on the offset calibration reference parameters.

[0006] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the adaptive calibration method for the measuring box described in any of the preceding claims.

[0007] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the adaptive calibration method for the measuring box described in any of the preceding claims.

[0008] The beneficial effects of this invention are as follows: It collects the operating noise of the metering box itself (including electromagnetic noise, background noise of electronic components, and micro-vibration of mechanical structures) under conditions without external load or business interference, forming a unique self-operating noise feature set, thus characterizing the "intrinsic characteristics" of the metering box from its source. Different metering boxes differ in component parameters, assembly processes, and aging levels. The self-operating noise feature set provides a personalized benchmark for each metering box, which helps avoid systematic errors caused by using a uniform model in subsequent calibration processes. Initializing and calibrating the noise feature set allows the metering model to directly match the actual operating state of the current metering box, rather than relying on factory calibration values, thereby reducing initial metering deviations. Outputting multiple metering data sequences (such as voltage, current, active power, reactive power, harmonic-related parameters, etc.) provides a sufficient data foundation for subsequent feature analysis and difference calculation, enhancing the information redundancy and robustness of the overall method. Through differential calculation between adjacent time periods, the changing characteristics of metering data within a short time scale can be amplified, enabling the system to quickly capture dynamic operating condition information such as load changes and operating state switching. Differential analysis inherently possesses the ability to eliminate static biases and slow-changing trends, helping to mitigate cumulative errors caused by long-term aging or slow temperature rise, allowing the focus of detection to be on real abnormal changes. Correlation analysis between environmental monitoring parameters such as temperature, humidity, electromagnetic field strength, and vibration and characteristic difference value sequences can clarify whether metrological anomalies are caused by changes in the external environment. Through source tracing analysis, different interference sources can be distinguished (such as electron drift caused by high temperature, signal distortion caused by electromagnetic interference, and transient noise caused by mechanical vibration), avoiding misjudging environmental factors as metrological faults. Based on the marked interference sources, corresponding offset calibration reference parameters are generated, giving the calibration process a clear direction. The offset calibration parameters are applied in real-time to operational metrological testing, forming a closed-loop control mechanism of "detection—analysis—calibration—re-detection," enabling the metrology box to have continuous self-adaptive capabilities. In complex and dynamically changing operating environments, this step can continuously correct metrological offsets caused by the environment and operating conditions, ensuring that metrological results maintain high accuracy and high consistency throughout their entire lifecycle. Attached Figure Description

[0009] Figure 1 This is a schematic diagram of the steps of an adaptive calibration method for a metering box according to the present invention; Figure 2 This is a detailed flowchart illustrating the implementation steps of step S1. Figure 3 This is a flowchart illustrating the detailed implementation steps of step S2. Detailed Implementation

[0010] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.

[0011] This application provides an adaptive calibration method, apparatus, device, and storage medium for a metering box. The execution entities of the adaptive calibration method, apparatus, device, and storage medium for the metering box include, but are not limited to, mechanical equipment, data processing platforms, cloud server nodes, and network upload devices that can be considered as general computing nodes in this application. The data processing platform includes, but is not limited to, at least one of an audio / image management system, an information management system, and a cloud data management system.

[0012] Please see Figures 1 to 3 This invention provides an adaptive calibration method for a metrology box, comprising the following steps: Step S1: Power on the metering box normally, perform self-operation noise detection, and construct a self-operation noise feature set; Step S2: Perform initial calibration and real-time measurement detection on the metering box according to the self-running noise feature set, and output multiple measurement data sequences; Step S3: Perform sequence-by-sequence feature analysis and difference calculation between adjacent time periods based on the multiple measurement data sequences to obtain a sequence of feature difference values; Step S4: Synchronously collect environmental monitoring parameters of the metering box; trace the source of environmental interference based on the characteristic difference value sequence according to the environmental monitoring parameters, and mark the interference source; Step S5: Perform offset calibration based on the interference source to obtain offset calibration reference parameters; perform real-time operating condition measurement and detection based on the offset calibration reference parameters.

[0013] In the embodiments of the present invention, see Figure 1 The diagram below illustrates the steps of an adaptive calibration method for a measuring box according to the present invention. In this example, the steps of the adaptive calibration method for the measuring box include: Step S1: Power on the metering box normally, perform self-operation noise detection, and construct a self-operation noise feature set; In this embodiment, the metering box is powered on under rated operating voltage and frequency conditions, allowing the internal metering chip, analog front-end, and sampling channel to enter a stable operating state. To ensure that the acquisition results only reflect the metering box's own operating characteristics, the metering input terminal is kept without effective load or in a high-impedance isolation state, so that the signal acquired by the sampling channel mainly consists of internal background noise. Self-operating noise detection is performed by directly acquiring discrete data from the original sampling channel output of the metering chip. Sampling parameters are set according to the characteristics of the metering chip; for example, the sampling frequency is selected as 4 kHz or 8 kHz, the sampling accuracy is 16 bits, and the continuous sampling duration is set to more than 60 seconds to ensure that the noise statistical results are sufficiently representative. The acquired noise data is preprocessed, including eliminating DC bias, removing abnormal abrupt changes, and limiting the analysis frequency band, thereby obtaining a stable self-operating noise sequence. Based on this, frequency and time domain analysis is performed on the noise sequence to extract characteristic parameters such as the dominant frequency distribution range, noise power spectral density, instantaneous amplitude variation range, and amplitude stability. Frequency stability and amplitude stability indices are obtained through time-series analysis.

[0014] Step S2: Perform initial calibration and real-time measurement detection on the metering box according to the self-running noise feature set, and output multiple measurement data sequences; In this embodiment, information such as frequency shift and amplitude stability changes reflected in noise characteristics are mapped to the metering parameter correction process. Initial adjustments are made to the gain coefficient, zero-point offset, and phase compensation parameters within the metering chip to ensure the metering channel reaches its optimal calibration state under its current operating conditions. After initial calibration, the metering box enters normal metering operation mode, performing real-time sampling and metering calculations on the received actual voltage and current signals. Metering data is generated according to a preset output cycle, such as outputting instantaneous quantities, power quantities, and energy data in units of 100 ms, 200 ms, or 1 s. All real-time metering results are corrected based on the initial calibration parameters to ensure the consistency and accuracy of the output data. Simultaneously, continuously output metering data is time-stamped and segmented according to a fixed time period, such as 10 s or 30 s as a time unit, merging metering data within the same time unit into a single metering data sequence.

[0015] Step S3: Perform sequence-by-sequence feature analysis and difference calculation between adjacent time periods based on the multiple measurement data sequences to obtain a sequence of feature difference values; In this embodiment, transient analysis is performed on the current data in each metering data sequence to calculate the current change rate, frequency of sudden changes, and duration, in order to characterize load changes and impact characteristics. Simultaneously, amplitude and phase calculations are performed on the voltage data to form voltage amplitude fluctuation trajectories and phase drift paths. Harmonic components within the sequence are identified, and the energy proportion of different harmonic components is calculated. After completing the above feature extraction, various features are processed to unify dimensions, constructing an energy feature matrix with time series as rows and feature parameters as columns. Based on this, difference calculations are performed on adjacent rows in the energy feature matrix according to time order to obtain the difference results such as feature change amplitude, change direction, and change rate.

[0016] Step S4: Synchronously collect environmental monitoring parameters of the metering box; trace the source of environmental interference based on the characteristic difference value sequence according to the environmental monitoring parameters, and mark the interference source; In this embodiment, while analyzing changes in metrological characteristics, environmental monitoring parameters related to the operation of the metering box are simultaneously collected, including the temperature and humidity gradients inside the box, changes in electromagnetic noise intensity, and the trend of changes in terminal contact resistance. Various environmental parameters are continuously sampled according to a unified time reference, for example, with a sampling period set to 1 second or 5 seconds, generating environmental parameter curves that change over time. Abrupt change detection is performed on the environmental parameter curves to identify and mark the time points where significant changes in environmental conditions occur. Subsequently, the marked environmental abrupt change time points are time-correlatedly matched with the feature difference value sequence to analyze whether there is a significant response in the differences in metrological characteristics before and after the environmental change. Through comprehensive analysis of the relationship between environmental parameter types, change amplitudes, and metrological characteristic change patterns, the impact path of environmental changes on metrological behavior is determined, thereby completing the source tracing of environmental interference and clearly marking the type of interference source, such as thermal environmental interference, electromagnetic interference, or contact anomaly interference.

[0017] Step S5: Perform offset calibration based on the interference source to obtain offset calibration reference parameters; perform real-time operating condition measurement and detection based on the offset calibration reference parameters.

[0018] In this embodiment, the interference intensity and direction of influence are determined based on the magnitude of changes in environmental parameters corresponding to the interference source and the resulting shift in metrological characteristics. A compensation amount opposite to the direction of the interference is then calculated. This compensation amount is mapped to the metrological parameter correction process to directionally adjust relevant parameters of the metrological channel, thereby offsetting the shift in metrological results caused by the interference. After parameter correction, the updated parameter combination is defined as the offset calibration reference parameter. During subsequent operation, the metrological box uses this offset calibration reference parameter to perform real-time metrological testing under operating conditions, ensuring that the metrological results continuously adapt to changes in current environmental conditions and operating status.

[0019] In this embodiment, see Figure 2 The diagram below illustrates the detailed implementation steps of step S1. In this embodiment, the detailed implementation steps of step S1 include: The metering box is powered on normally, and the self-running noise is detected and extracted based on the original sampling channel of the metering chip. Based on the self-running noise, the main frequency distribution range, noise power spectral density, instantaneous amplitude variation range, and amplitude stability index are calculated to obtain the self-noise characteristics; Time-series drift analysis was performed on the self-noise characteristics to obtain frequency stability index and amplitude stability index; A self-running noise feature set is constructed based on frequency stability index and amplitude stability index.

[0020] In this embodiment, the power metering box is powered on under rated operating conditions, allowing the metering chip to enter normal operating mode. To ensure that the collected data only reflects the operating characteristics of the metering box itself, no effective measurement signal is connected to the metering input terminal, and the voltage and current sampling channels are in an unloaded or high-impedance isolated state. At this time, the data collected by the original sampling channel of the metering chip mainly comes from the background noise of the analog front-end device, the quantization noise during the analog-to-digital conversion process, and the inherent disturbances introduced by the chip's internal reference source and clock circuit. By directly reading the unfiltered and uncalculated raw sampling data of the metering chip, a self-operating noise acquisition channel is constructed. The sampling parameters are set according to the operating characteristics of the metering chip, for example, the sampling frequency is selected as 4 kHz or 8 kHz, the sampling accuracy is 16 bits, and the continuous sampling duration is set to more than 60 seconds to cover multiple power grid cycles and improve statistical stability. During the sampling process, the raw data undergoes basic processing, including eliminating DC bias, removing sudden abnormal points exceeding reasonable thresholds, and suppressing extremely low frequency drift and high frequency random glitches by limiting the frequency band. Noise data is segmented according to fixed time windows, such as 1 second per analysis window, and partial overlap is used to improve the continuity of analysis. For each segment of noise data, a frequency domain transformation is performed to obtain the corresponding spectral distribution, and statistical analysis is performed on the spectral energy to determine the frequency range where the energy proportion is concentrated. This range is defined as the noise dominant frequency distribution interval to characterize the frequency characteristics corresponding to the main noise sources inside the metering chip. Based on this, the noise power spectral density distribution is calculated by averaging multiple spectral results to quantify the change of noise energy with frequency in different frequency bands. At the same time, statistical analysis is performed on the noise signal in the time domain to calculate the instantaneous maximum, minimum, and peak-to-peak values ​​within each analysis window to describe the range of noise amplitude variation. Amplitude stability-related indicators are also introduced, such as calculating the dispersion of the noise root mean square value within multiple time windows, to measure the consistency of noise amplitude changes over time.

[0021] The center frequency, bandwidth, and corresponding energy percentage of the main frequency distribution range are used as frequency characteristic parameters and recorded at fixed time intervals to form a time series of frequency characteristics. By performing sliding window statistical analysis or trend fitting on this series, the fluctuation range, slope, and dispersion of the frequency characteristics in the time dimension are calculated, and a frequency stability index is formed to characterize the stable performance of the internal oscillation source and analog link of the metering chip during operation. Simultaneously, noise amplitude-related characteristics, such as root mean square (RMS) and peak value, are analyzed using the same time series analysis method to statistically determine their variation amplitude and consistency level over a long time scale, thus obtaining an amplitude stability index. Different indices are standardized in terms of dimensions; for example, frequency drift amplitude, power spectrum fluctuation, and amplitude stability parameters are mapped to a unified numerical range through normalization to avoid the influence of different dimensions on subsequent analysis. Subsequently, based on the sensitivity of metering accuracy and the requirements for operational status assessment, various indices are weighted and combined to ensure that frequency stability and amplitude stability form a complementary relationship in the feature set. The final self-operating noise feature set exists in the form of a multi-dimensional parameter vector, which can comprehensively reflect the inherent operating characteristics of the metering box under current dynamic conditions.

[0022] In this embodiment, see Figure 3 The diagram below illustrates the detailed implementation steps of step S2. In this embodiment, the detailed implementation steps of step S2 include: The metering box is initialized and calibrated based on the self-operating noise feature set to obtain the initial calibration reference. The metering box is driven to operate normally according to the initial calibration benchmark, and real-time metering detection is performed to output real-time metering data. The real-time metering data is timestamped and synchronized to extract the time-series data stream; Set a fixed time period; The time-series data stream is divided and sequentially numbered according to the fixed time period, and multiple measurement data sequences are output.

[0023] In this embodiment, the main frequency distribution range, noise power spectral density characteristics, frequency stability index, and amplitude stability index contained in the self-operating noise feature set are used as calibration reference inputs to evaluate the current background offset state of the metering channel. By comparing with a pre-set reference noise threshold range, the deviation of each feature parameter from the ideal stable state is analyzed, and the gain coefficient, phase compensation parameter, and zero-point offset of the metering chip are initially corrected accordingly. The adjustment range of the relevant parameters is mapped according to the proportion of noise feature change. For example, when the main frequency center shifts slightly, the internal time base compensation factor is adjusted accordingly; when the amplitude stability index is too high, the analog front-end gain is slightly reduced. The calibration process is completed without introducing external measurement signals, ensuring that the calibration results only reflect the operating characteristics of the metering box itself. After completing the initial calibration and obtaining the initial calibration reference, the metering box enters the normal operation state according to the corrected parameter configuration and begins to perform real-time metering processing on the actual input voltage and current signals. At this time, the metering chip continuously samples the input signal at the rated sampling frequency (e.g., 4 kHz or 8 kHz) and completes data correction according to the gain, phase, and zero-point compensation parameters in the initial calibration reference. During real-time metering and testing, continuous output of metering data includes instantaneous voltage, instantaneous current, active power, reactive power, and cumulative energy. To ensure data continuity and stability, the output period of the metering data can be set to 100 ms, 200 ms, or 1 s, depending on the application scenario. All output data are calculated based on a unified calibration benchmark, ensuring high consistency and accuracy of the metering results in the initial stage.

[0024] A unified time reference source is established within the metering box, assigning a unique time identifier to each metering data generation moment. The time accuracy can be set to milliseconds or higher to meet the needs of subsequent time-series analysis. The introduction of timestamps ensures that metering data such as voltage, current, and power not only contain numerical information but also clear information about the order of occurrence and time intervals. Subsequently, the timestamped metering data is organized according to the generation order, forming a continuous time-series data stream structure. This data stream can accurately reflect the dynamic changes of the metering box during operation, providing a foundation for subsequent data segmentation according to fixed time periods.

[0025] A unified constraint is imposed on the time scale of data analysis by setting a fixed time period to standardize subsequent data partitioning. The fixed time period is selected based on the metrological accuracy requirements and the rate of change in operating status, such as commonly used time lengths like 5 s, 10 s, 30 s, or 60 s. A time period that is too short may lead to excessive data fluctuations, hindering the extraction of stable features; a time period that is too long may mask short-term state changes. Therefore, in practical applications, a period length that balances response speed and statistical stability is usually chosen. This fixed time period remains consistent throughout the entire operation, serving as a unified reference scale for time-series data partitioning and numbering. Based on timestamps, metrological data within the same time period are grouped into a single metrological data sequence, with each sequence containing all real-time metrological data points within that period. After partitioning, each metrological data sequence is sequentially numbered according to its chronological order, for example, labeled as Period 1, Period 2, and so on, thus forming multiple sets of metrological data sequences with a clear structure and explicit order. Each set of data sequences represents the operating and metrological status of the metering box within its corresponding time period. In this way, continuous real-time data can be transformed into multiple discrete but time-correlated data units, which facilitates subsequent comparative analysis of measurement changes, noise effects, and calibration offsets between different periods.

[0026] In this embodiment, step S3 includes the following steps: Based on the multiple measurement data sequences, sequential transient current analysis is performed to calculate the current change rate, frequency of sudden changes and duration, thus forming current transient characteristics. Voltage amplitude and phase are calculated for the multiple metering data sequences, and voltage amplitude fluctuation trajectory and phase drift path are extracted; Harmonic signals are identified from the multiple metering data sequences, and the energy percentage is calculated to obtain the harmonic capability percentage of different sequences. The transient characteristics of the current, the proportion of harmonic capability, the voltage amplitude fluctuation trajectory, and the phase drift path are processed with unified dimensions to construct multiple sequences of power characteristic matrices. The electrical energy feature matrix is ​​subjected to adjacent time period difference calculation to obtain a feature difference value sequence.

[0027] In this embodiment, transient current characteristic analysis is performed on each sequence. Each metering data sequence corresponds to continuous current sampling data within a fixed time period, with a consistent sampling interval, such as 0.25 ms or 0.5 ms. First, the instantaneous current data within the sequence is processed using first-order difference to calculate the current change between adjacent sampling points. This current change rate is then combined with the sampling time interval to reflect the drastic changes in load status over a short period. Subsequently, by setting a change rate threshold, change points exceeding the threshold are identified and judged as current abrupt events. Based on this, the number of abrupt events within each metering data sequence is counted, and the duration from the occurrence of each abrupt event to its return to stability is recorded, thereby quantifying the frequency and duration characteristics of transient disturbances. To avoid the influence of occasional noise on the analysis results, a minimum duration constraint is typically introduced, for example, only change events with a duration exceeding 2 ms are statistically analyzed. The effective value or periodic mean of the instantaneous voltage is extracted using an amplitude calculation method, and a continuous voltage amplitude change sequence is formed according to the sampling order to describe the voltage fluctuation trajectory over time. This trajectory can intuitively reflect the stability of power supply conditions over a short time scale. Subsequently, phase calculations are performed on the voltage signal. By extracting the phase angle of the fundamental component, the phase change within the corresponding time period is obtained, and the continuous phase results are connected in chronological order to form a phase drift path. To improve the stability of the phase calculation, one or more fundamental periods are typically used as the analysis window to avoid the influence of instantaneous distortion on phase judgment. By simultaneously obtaining the voltage amplitude fluctuation trajectory and the phase drift path, the variation characteristics of the voltage state within each measurement data sequence can be comprehensively described from both amplitude and phase dimensions.

[0028] The voltage or current sampling data within the sequence is converted to the frequency domain, and the fundamental frequency and its integer multiples are identified through spectral analysis. For common applications, the 2nd to 25th harmonic components are the primary focus, and the energy value corresponding to each harmonic is calculated. Subsequently, the ratio of each harmonic energy to the total signal energy within the sequence is calculated to obtain the energy proportion of different harmonic components, which are then summarized to form the harmonic capability proportion index of the measurement data sequence. To reduce the impact of spectral leakage on the results, windowing is typically applied to the data during the analysis, and results from multiple periods are averaged. Various features are uniformly organized and their dimensions are processed to construct an energy characteristic matrix suitable for comprehensive analysis. Since the current rate of change, frequency of abrupt changes, harmonic energy proportion, and voltage amplitude and phase parameters differ in numerical range and physical meaning, various features are first normalized, for example, mapped to a numerical range of 0 to 1, to eliminate the influence of dimensional differences. Then, a two-dimensional energy characteristic matrix is ​​constructed with the measurement data sequence as rows and various characteristic parameters as columns, in a unified order. Each row in the matrix fully describes the electrical state characteristics within the corresponding time period, including the degree of transient impact, voltage stability, and harmonic content.

[0029] By performing differential calculations on adjacent time periods, the differences in electrical energy characteristics over time are extracted. Specifically, the eigenvectors corresponding to adjacent rows in the electrical energy feature matrix are differentially analyzed item by item to calculate the magnitude of change of each feature parameter over consecutive time periods. This differential result quantifies the degree of change in transient current characteristics, voltage fluctuation characteristics, and harmonic proportions over time. Subsequently, each set of adjacent differential results is arranged in chronological order to form a sequence of feature difference values. This sequence highlights time points where significant changes in operating status occur while mitigating minor fluctuations under long-term stable conditions.

[0030] In this embodiment, the specific steps for performing adjacent time-time difference calculations on the electrical energy feature matrix to obtain the feature difference value sequence are as follows: The electrical energy characteristic matrix is ​​subjected to differential calculation of adjacent time periods to obtain the differential results of adjacent segments; the differential results include the characteristic change amplitude, change direction and change rate; Trend analysis is performed on the difference results to obtain difference trend data; Perform time-series difference statistics on the differential trend data to obtain a sequence of characteristic difference values.

[0031] In this embodiment, adjacent rows of feature vectors in the matrix are selected in chronological order. Each corresponding feature dimension is compared item by item, and the change in value between the previous and subsequent time periods is calculated. This change is the feature change amplitude, used to quantify the degree of change of the feature parameter within adjacent time periods. Subsequently, based on the positive or negative relationship of the change, the direction of feature change is determined to distinguish whether the feature parameter is in an upward trend, a downward trend, or remains essentially unchanged. Simultaneously, combined with the length of a fixed time period, such as 10 s or 30 s, the feature change amplitude is divided by the time period length to obtain the rate of change per unit time, reflecting the speed of feature change. The difference results of multiple consecutive time periods are arranged in chronological order to construct a sequence of difference parameters changing over time. For each type of feature parameter, its change amplitude, direction, and rate of change are jointly analyzed. Using sliding window statistics or multi-period cumulative analysis, it is identified whether the feature change exhibits a trend of continuous enhancement, continuous weakening, or periodic fluctuation. For example, when a feature maintains a consistent direction of change and its rate of change gradually increases over multiple consecutive time periods, it can be determined that the feature is in a state of obvious trend change; when the direction of change frequently reverses and the magnitude of change is small, it can be determined that the feature is in a relatively stable state. To avoid the influence of short-term occasional disturbances on trend judgment, a minimum trend judgment period is usually set, such as three or five consecutive time periods meeting the condition of the same direction of change before it is considered a valid trend.

[0032] Differential trend data are grouped and statistically analyzed in chronological order. Trend characteristics within each time period are quantified and summarized, such as the cumulative value of trend change magnitude, trend duration, and average rate of change. By standardizing the dimensions of these statistical results, the trend differences of different characteristics are mapped to comparable numerical ranges. Subsequently, the statistical results corresponding to each time period are arranged in chronological order to form a continuous sequence of characteristic difference values. This sequence can intuitively reflect the degree and intensity of differences in electrical characteristics across different time periods, highlighting time points where significant changes in operating status occur, while mitigating the impact of short-term random fluctuations.

[0033] In this embodiment, the specific steps of step S4 are as follows: Simultaneously collect environmental monitoring parameters of the metering box; the environmental monitoring parameters include the temperature and humidity change gradient inside the box, the change in electromagnetic noise intensity, and the change trend of the contact resistance of the wiring terminals. The environmental monitoring parameters are continuously sampled to generate environmental parameter curves; Perform abrupt change detection on the environmental parameter curves and mark the abrupt change time points; Based on the mutation time points, the characteristic difference value sequences are correlated and matched to identify the corresponding measurement change information; Environmental interference sources are traced and marked based on corresponding measurement change information.

[0034] In this embodiment, the environmental conditions closely related to the metering stability during the operation of the metering box are simultaneously sensed and parameters are collected. Environmental monitoring parameters mainly include the temperature and humidity gradients inside the box, changes in electromagnetic noise intensity, and the trend of contact resistance at the terminals. Specifically, temperature and humidity sensing units are deployed inside the metering box and at key locations to collect temperature and humidity data at different locations within the box. The temperature and humidity gradients are calculated by the difference between adjacent sampling points, thus reflecting the dynamic changes in the internal thermal environment. Simultaneously, electromagnetic noise intensity around and inside the metering box is collected using electromagnetic sensing units, with a focus on the amplitude of changes in low-frequency interference and high-frequency radiation to characterize the potential impact of the external electromagnetic environment on metering behavior. Furthermore, the trend of changes in equivalent contact resistance is calculated by indirectly measuring the voltage and current characteristics at the terminals, reflecting changes in connection status caused by factors such as terminal loosening, oxidation, or aging. The temperature, humidity, electromagnetic noise intensity, and contact resistance changes are arranged chronologically to construct corresponding time-series data. To improve the smoothness and analyzability of the curves, the raw sampling data is usually smoothed appropriately, for example, by using a moving average to eliminate the impact of instantaneous jitter on the overall trend judgment. The environmental parameter curves, with fixed time intervals on the horizontal axis and corresponding parameter values ​​on the vertical axis, can intuitively reflect the changing trajectory of the environment inside and around the enclosure during operation.

[0035] The changes between adjacent sampling points are calculated for various environmental parameter curves, and abrupt change thresholds are set based on the normal fluctuation range of the parameters themselves. For example, when the temperature or humidity gradient exceeds the set threshold within a short period, it can be identified as a thermal environment abrupt change; when the electromagnetic noise intensity shows a significant jump within a certain time period, it can be identified as an electromagnetic interference abrupt change; when the contact resistance trend shifts significantly within a short period, it can be identified as an abnormal connection status. To avoid misjudgments caused by occasional noise, the abrupt change status is usually required to remain consistent across multiple consecutive sampling points before confirmation. Centered on the abrupt change time point, the characteristic changes within the corresponding or adjacent time periods are searched in the characteristic difference value sequence to analyze whether there are significant metrological characteristic difference responses. By comparing the magnitude and direction of the characteristic difference values ​​before and after the abrupt change, the influence pattern of environmental changes on metrological behavior can be identified. For example, when a significant jump or trend change appears in the characteristic difference value sequence after an environmental abrupt change, it can be determined that there is a correlation between the environmental change and the metrological change.

[0036] In this embodiment, the specific steps of step S5 are as follows: Determine the interference intensity and direction of influence of the interference source; Based on the interference intensity and direction of influence, reverse interference compensation calculation is performed to obtain the compensation value; The offset calibration of the metering box is performed based on the compensation value to obtain the offset calibration reference parameters. Real-time operating condition metrology testing is performed based on offset calibration reference parameters.

[0037] In this embodiment, the magnitude and duration of sudden changes in environmental parameters, along with the corresponding differences in metrological characteristics, are jointly analyzed. By establishing a correspondence between environmental changes and metrological offsets, the impact of interference sources on metrological behavior is assessed. Interference intensity can be calculated by normalizing the changes in environmental parameters and then combining this with the magnitude of changes in metrological characteristic differences. For example, if the electromagnetic noise intensity shows a significant jump within a short period, and the metrological characteristic difference increases synchronously, the interference source can be determined to have high interference intensity. The direction of influence is determined by analyzing the sign and trend of changes in metrological parameters, such as directional characteristics like high or low power values, or phase advance or lag. Based on the mechanisms of different types of interference sources, a mapping relationship between interference influence and metrological offset is constructed, and the sign of the compensation parameters is set according to the direction of influence, ensuring that the compensation effect is opposite to the direction of interference influence. For example, when interference causes the overall metrological results to be too high, the compensation value is set as a negative correction; when interference causes phase lag, the compensation value is set as a positive phase adjustment. The magnitude of the compensation value is proportionally mapped to the interference intensity; the greater the interference intensity, the greater the corresponding compensation magnitude. To avoid overcompensation causing new deviations, a maximum adjustment limit is usually set for the compensation value, and compensation is applied gradually using a smooth transition method.

[0038] The compensation value is mapped to the corresponding calibration parameters of the metering channel, such as gain correction coefficient, zero offset, or phase compensation parameters, and the current parameter configuration of the metering box is updated according to predetermined rules. During calibration, the metering box is kept running continuously, and compensation correction is introduced gradually to ensure smooth changes in measurement results and avoid new instability factors caused by sudden parameter changes. After calibration, the updated parameter set is saved as the offset calibration reference parameter. This parameter set fully reflects the compensation state required to maintain measurement accuracy under current environmental interference conditions. While continuously collecting basic measurement data such as voltage and current, the offset calibration reference parameter is called in real time to correct the sampled data, enabling the measurement results to dynamically offset the offset effects caused by current environmental interference. As the operation progresses, when environmental conditions or interference conditions change, the measurement detection process can be continuously updated with new offset calibration results, thus forming a closed-loop adaptive calibration mechanism. In this way, the metering box can maintain the stability and consistency of measurement results under different operating conditions, ensuring that real-time operating condition measurement is always based on a calibration reference that matches the current environmental conditions, ultimately achieving high-reliability measurement and continuous accuracy maintenance capabilities based on dynamic operating condition awareness.

[0039] In this embodiment, an adaptive calibration device for a metrology box is provided, used to perform the adaptive calibration method for a metrology box as described above, including: The noise detection unit is used to drive the metering box to power on normally, perform self-operation noise detection, and construct a self-operation noise feature set. The metering unit is used to perform initial calibration and real-time metering detection on the metering box according to the self-operating noise feature set, and output multiple metering data sequences. The difference calculation unit is used to perform sequence feature analysis and difference calculation between adjacent time periods based on the multiple measurement data sequences to obtain a sequence of feature difference values. The interference tracing unit is used to synchronously collect environmental monitoring parameters from the metering box; and to trace the source of environmental interference based on the characteristic difference value sequence according to the environmental monitoring parameters, and mark the interference source. The calibration unit is used to perform offset calibration based on the interference source to obtain offset calibration reference parameters; and to perform real-time operating condition measurement and detection based on the offset calibration reference parameters.

[0040] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the adaptive calibration method for the measuring box described in any of the preceding claims.

[0041] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the adaptive calibration method for the measuring box described in any of the preceding claims.

[0042] Therefore, the embodiments should be considered as exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of the equivalents of the application are intended to be included within the invention.

[0043] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement it. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein are implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. An adaptive calibration method for a measuring chamber, characterized in that, Includes the following steps: Step S1: Power on the metering box normally, perform self-operation noise detection, and construct a self-operation noise feature set; Step S2: Perform initial calibration and real-time measurement detection on the metering box according to the self-running noise feature set, and output multiple measurement data sequences; Step S3: Perform sequence-by-sequence feature analysis and difference calculation between adjacent time periods based on the multiple measurement data sequences to obtain a sequence of feature difference values; Step S4: Synchronously collect environmental monitoring parameters of the metering box; trace the source of environmental interference based on the characteristic difference value sequence according to the environmental monitoring parameters, and mark the interference source; Step S5: Perform offset calibration based on the interference source to obtain offset calibration reference parameters; perform real-time operating condition measurement and detection based on the offset calibration reference parameters.

2. The adaptive calibration method for the measuring box according to claim 1, characterized in that, The specific steps of step S1 are as follows: The metering box is powered on normally, and the self-running noise is detected and extracted based on the original sampling channel of the metering chip. Based on the self-running noise, the main frequency distribution range, noise power spectral density, instantaneous amplitude variation range, and amplitude stability index are calculated to obtain the self-noise characteristics; Time-series drift analysis was performed on the self-noise characteristics to obtain frequency stability index and amplitude stability index; A self-running noise feature set is constructed based on frequency stability index and amplitude stability index.

3. The adaptive calibration method for the measuring box according to claim 1, characterized in that, The specific steps of step S2 are as follows: The metering box is initialized and calibrated based on the self-operating noise feature set to obtain the initial calibration reference. The metering box is driven to operate normally according to the initial calibration benchmark, and real-time metering detection is performed to output real-time metering data. The real-time metering data is timestamped and synchronized to extract the time-series data stream; Set a fixed time period; The time-series data stream is divided and sequentially numbered according to the fixed time period, and multiple measurement data sequences are output.

4. The adaptive calibration method for the measuring box according to claim 1, characterized in that, Step S3 is as follows: Based on the multiple measurement data sequences, sequential transient current analysis is performed to calculate the current change rate, frequency of sudden changes and duration, thus forming current transient characteristics. Voltage amplitude and phase are calculated for the multiple metering data sequences, and voltage amplitude fluctuation trajectory and phase drift path are extracted; Harmonic signals are identified from the multiple metering data sequences, and the energy percentage is calculated to obtain the harmonic capability percentage of different sequences. The transient characteristics of the current, the proportion of harmonic capability, the voltage amplitude fluctuation trajectory, and the phase drift path are processed with unified dimensions to construct multiple sequences of power characteristic matrices; The electrical energy feature matrix is ​​subjected to adjacent time period difference calculation to obtain a feature difference value sequence.

5. The adaptive calibration method for the measuring box according to claim 4, characterized in that, The specific steps for performing adjacent time-time difference calculations on the electrical energy feature matrix to obtain the feature difference value sequence are as follows: The electrical energy characteristic matrix is ​​subjected to differential calculation of adjacent time periods to obtain the differential results of adjacent segments; the differential results include the characteristic change amplitude, change direction and change rate; Trend analysis is performed on the difference results to obtain difference trend data; Perform time-series difference statistics on the differential trend data to obtain a sequence of characteristic difference values.

6. The adaptive calibration method for the measuring box according to claim 1, characterized in that, The specific steps of step S4 are as follows: Simultaneously collect environmental monitoring parameters of the metering box; the environmental monitoring parameters include the temperature and humidity change gradient inside the box, the change in electromagnetic noise intensity, and the change trend of the contact resistance of the wiring terminals. The environmental monitoring parameters are continuously sampled to generate environmental parameter curves; Perform abrupt change detection on the environmental parameter curves and mark the abrupt change time points; Based on the mutation time points, the characteristic difference value sequences are correlated and matched to identify the corresponding measurement change information; Environmental interference sources are traced and marked based on corresponding measurement change information.

7. The adaptive calibration method for the measuring box according to claim 1, characterized in that, The specific steps of step S5 are as follows: Determine the interference intensity and direction of influence of the interference source; Based on the interference intensity and direction of influence, reverse interference compensation calculation is performed to obtain the compensation value; The offset calibration of the metering box is performed based on the compensation value to obtain the offset calibration reference parameters. Real-time operating condition metrology testing is performed based on offset calibration reference parameters.

8. An adaptive calibration device for a metering box, characterized in that, An adaptive calibration method for a metrology box as described in claim 1 includes: The noise detection unit is used to drive the metering box to power on normally, perform self-operation noise detection, and construct a self-operation noise feature set. The metering unit is used to perform initial calibration and real-time metering detection on the metering box according to the self-operating noise feature set, and output multiple metering data sequences. The difference calculation unit is used to perform sequence feature analysis and difference calculation between adjacent time periods based on the multiple measurement data sequences to obtain a sequence of feature difference values. The interference tracing unit is used to synchronously collect environmental monitoring parameters from the metering box; and to trace the source of environmental interference based on the characteristic difference value sequence according to the environmental monitoring parameters, and mark the interference source. The calibration unit is used to perform offset calibration based on the interference source to obtain offset calibration reference parameters; and to perform real-time operating condition measurement and detection based on the offset calibration reference parameters.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the adaptive calibration method for the metrology box according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the adaptive calibration method for the metrology box according to any one of claims 1 to 7.

Images