Metering data automatic calibration method and system

Abstract

The invention provides a metering data automatic calibration method and system, and the method comprises the steps: obtaining an actual measurement output value of a target intelligent sensor for a measured physical quantity in a normal operation process of metering equipment, accessing a standard reference signal outputted by an external standard device subjected to the traceability of the quantity value, reading a pre-stored standard reference value and a standard threshold value, and carrying out the calibration of the measured physical quantity. According to the actually measured output value, the standard reference value and the standard reference signal, obtaining a standardized output value through a fitting mode; when the standardized output value exceeds a standard threshold value, switching to a calibration mode, providing forward and reverse standard input quantities by an external standard device, acquiring a response signal, and calculating a nonlinear compensation coefficient and a hysteresis correction coefficient; and updating the output correction parameter according to the compensation coefficient. By adopting the scheme of the invention, the cost constraint of different error attention in measurement data calibration can be realized, and the calibration robustness is improved.

Description

Technical Field

[0001] This application relates to the field of metrology and testing technology, and more specifically, to an automatic calibration method and system for metrological data. Background Technology

[0002] In the management of metrology equipment, metrology data calibration is a fundamental step in ensuring measurement accuracy. It is of great significance for industrial process control, product quality inspection and scientific research optimization. Traditional metrology calibration methods mostly focus on periodic shutdown calibration and manual intervention, lacking the ability to dynamically identify and adaptively correct real-time response deviations of sensors. This makes it difficult to support the continuous and reliable output of metrology data under complex operating conditions. With the increasing intelligence of metrology equipment and the increasing complexity of application scenarios, real-time online calibration and high-precision error compensation capabilities have gradually become the core requirements of metrology systems.

[0003] Current mainstream metrological data calibration methods generally suffer from static design issues, making it difficult to adapt to nonlinear and hysteresis biases caused by environmental interference or aging during sensor operation. This is especially true in scenarios with frequent fluctuations in physical quantities or frequent sensor response anomalies, where the nonlinear correlation between measured output values ​​and standard reference values ​​is difficult to extract effectively. This results in insufficient calibration responsiveness to different error types. The root cause of this deficiency lies in the fact that traditional methods fail to introduce compensation mechanisms that reflect real-time deviation distribution, lack independent suppression measures for nonlinear and hysteresis biases, and lack strategies to enhance key response characteristics. Consequently, this limits the calibration's analytical capability and stability in terms of metrological accuracy. Therefore, how to achieve cost constraints on different error attention levels in metrological data calibration, thereby improving the robustness of automatic metrological data calibration, has become a challenge for the industry. Summary of the Invention

[0004] This application provides an automatic calibration method and system for metrological data, which can realize the cost constraint of different error attention in metrological data calibration.

[0005] In a first aspect, this application provides an automatic calibration method for metrological data, comprising: During the normal operation of the metering equipment, the measured output value of the physical quantity being measured by the target intelligent sensor in the metering equipment is obtained, and at the same time, the standard reference signal output by the external standard that has been traced by the measurement value is connected. Read the standard reference value and standard threshold value pre-stored in the calibration module, and determine the standardized output value based on the measured output value and the standard reference value; The standardized output value is compared with the standard threshold. When the standardized output value exceeds the standard threshold of the target smart sensor, the metering device is controlled to switch to calibration mode. An external standard provides a positive standard input and a negative standard input with known values, and the first response signal and the second response signal corresponding to the target smart sensor are collected simultaneously. Based on the first response signal, the second response signal, and the corresponding positive and negative standard input quantities, the nonlinear compensation coefficient and hysteresis correction coefficient of the target intelligent sensor are calculated with reference to the standard reference value. The output correction parameters of the target smart sensor are updated based on the nonlinear compensation coefficient and the hysteresis correction coefficient.

[0006] Preferably, determining the standardized output value based on the measured output value and the standard reference value specifically includes: Multiple sets of measured output values ​​and standard reference values ​​are obtained from multiple sampling windows respectively. The measured output values ​​and the standard reference values ​​are time-aligned, and the correlation mapping features between the measured output values ​​and the standard reference signal within the same sampling window are extracted. The dynamic correlation coefficient between the measured output value and the standard reference value is determined based on multiple sets of measured output values ​​and multiple sets of standard reference values. The standardized output value is obtained by weighting and summing all the correlation mapping features using all the dynamic correlation coefficients.

[0007] Preferably, the standardized output value is compared with the standard threshold. When the standardized output value exceeds the standard threshold of the target smart sensor, the measurement device is controlled to switch to calibration mode. An external standard provides positive and negative standard input values ​​of known quantities. The first and second response signals corresponding to the target smart sensor are simultaneously acquired, specifically including: Compare the deviation between the measured output value and the standard reference value; When the deviation exceeds the preset threshold, the metering device is triggered to enter the calibration mode, and positive and negative standard input quantities are sequentially applied to the target smart sensor through an external standard. The first response signal is acquired when a positive standard input is applied, and the second response signal is acquired when a negative standard input is applied.

[0008] Preferably, the nonlinear compensation coefficient and hysteresis correction coefficient of the target intelligent sensor are calculated based on the first response signal, the second response signal, and the corresponding positive and negative standard input quantities, with reference to standard reference values. Specifically, this includes: Based on the positive and negative standard input values ​​at multiple different range points, multiple sets of first and second response signals are simultaneously acquired. Extract the deviation distribution characteristics between the first response signal and the positive standard input, and the deviation distribution characteristics between the second response signal and the negative standard input; The nonlinear response intensity and hysteresis response intensity are determined based on the aforementioned deviation distribution characteristics; An initial nonlinear compensation factor is set based on the nonlinear response intensity, and an initial hysteresis correction factor is set based on the hysteresis response intensity. The nonlinear compensation coefficient is obtained by calculating the error between the initial nonlinear compensation factor and the real-time standard value provided by the current external standard, and by repeatedly adjusting the initial nonlinear compensation factor to minimize the error; the hysteresis correction coefficient is obtained by calculating the error between the initial hysteresis correction factor and the real-time standard value provided by the current external standard, and by repeatedly adjusting the initial hysteresis correction factor to minimize the error.

[0009] Preferably, updating the output correction parameters of the target intelligent sensor based on the nonlinear compensation coefficient and the hysteresis correction coefficient specifically includes: The nonlinear compensation coefficient is applied to the output function of the target intelligent sensor to compensate for the nonlinear deviation. The hysteresis correction coefficient is applied to the output function of the target smart sensor to correct the hysteresis deviation; Based on the compensated and corrected output function, the value of the output correction parameter is adjusted by combining multiple sets of historical calibration data to adapt it to the current response characteristics of the sensor.

[0010] Preferably, reading the standard reference values ​​and standard thresholds pre-stored in the calibration module specifically includes: Obtain the standard reference value sequence for measurement traceability from the storage unit of the calibration module; The standard threshold is determined based on statistical analysis of historical measurement data, and the standard threshold is ensured to be suitable for the measurement range of the target smart sensor, wherein the standard reference value is the real-time standard physical quantity value provided by the current external standard.

[0011] Preferably, the external standard is a precision calibrator or reference metrology device that can provide standard physical quantities with traceable values.

[0012] Secondly, this application provides an automatic calibration system for metrological data, comprising: The acquisition module is used to acquire the measured output value of the target intelligent sensor of the metering equipment for the measured physical quantity during normal operation of the metering equipment, and at the same time connect to the standard reference signal output by the external standard instrument that has been traced by the measurement value. The processing module is used to read the standard reference value and standard threshold value pre-stored in the calibration module, and determine the standardized output value based on the measured output value and the standard reference value; The processing module is also used to compare the standardized output value with the standard threshold. When the standardized output value exceeds the standard threshold of the target smart sensor, the metering device is controlled to switch to calibration mode. An external standard provides a positive standard input and a negative standard input with known values, and the first response signal and the second response signal corresponding to the target smart sensor are collected simultaneously. The processing module is also used to calculate the nonlinear compensation coefficient and hysteresis correction coefficient of the target intelligent sensor based on the first response signal, the second response signal and the corresponding positive and negative standard input quantities, with reference to the standard reference value. The execution module is used to update the output correction parameters of the target smart sensor based on the nonlinear compensation coefficient and the hysteresis correction coefficient.

[0013] Thirdly, this application provides a computer device, the computer device including a memory and a processor, the memory storing code, and the processor being configured to acquire the code and execute the above-described automatic calibration method for metrological data.

[0014] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described automatic calibration method for metrological data.

[0015] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects: In this embodiment, during normal operation of the metrology equipment, the measured output value of the target intelligent sensor for the measured physical quantity is acquired, and a standard reference signal output by an external standard traceable by the metrology equipment is simultaneously connected. The standard reference value and standard threshold value pre-stored in the calibration module are read, and a standardized output value is determined based on the measured output value and the standard reference value. The standardized output value is compared with the standard threshold value. When the standardized output value exceeds the standard threshold value of the target intelligent sensor, the metrology equipment is controlled to switch to calibration mode. The external standard provides a positive standard input and a negative standard input of known quantities, and the first response signal and the second response signal corresponding to the target intelligent sensor are simultaneously acquired. Based on the first response signal, the second response signal, and the corresponding positive and negative standard inputs, the nonlinear compensation coefficient and hysteresis correction coefficient of the target intelligent sensor are calculated with reference to the standard reference value. The output correction parameters of the target intelligent sensor are updated according to the nonlinear compensation coefficient and the hysteresis correction coefficient.

[0016] Therefore, this application updates the output correction parameters of the target intelligent sensor through nonlinear compensation coefficients and hysteresis correction coefficients. Firstly, by performing time alignment and feature extraction on the measured output value and the standard reference signal, it can reveal the sensor's response to changes in physical quantities and its deviation characteristics, under the premise of traceability of measurement values. This establishes a calibration feature expression mechanism with error resolution capabilities. This fitting process extracts the coupling characteristics between measured values ​​and standard values, realizing multi-set correlation expressions of measurement data, and obtaining standardized output values ​​on this basis, thus enhancing the calibration's ability to perceive error differences from the source. Secondly, by determining the deviation between the standardized output value and the standard threshold, the calibration mode can be dynamically triggered, effectively suppressing the interference of abnormal deviations on the measurement results, highlighting the dominant characteristics of key errors, and realizing error resolution. Controllable adjustment of attention provides a clearer basis for subsequent compensation. Then, based on multiple sets of positive and negative input quantities, sensor nonlinearity and hysteresis biases are predicted. Combined with bias distribution characteristics, an error minimization method is used to quantitatively adjust the compensation coefficients, thereby outputting a compensation value with uncertainty tolerance, significantly improving the adaptability of calibration to complex error behavior. Finally, when the bias exceeds a threshold, a cost constraint mechanism activates the readjustment process of the output correction parameters, further strengthening the compensation expression of key errors in abnormal tasks and minimizing the misleading effect of noise bias on calibration results. This mechanism, through dynamic compensation, constructs an adaptive adjustment path for calibration in error fluctuation scenarios, effectively overcoming the problem of insufficient robustness caused by fixed error expression patterns in traditional calibration methods. In summary, the proposed solution can realize cost constraints on different error attention in metrological data calibration, thereby improving the robustness of automatic metrological data calibration. Attached Figure Description

[0017] Figure 1 This is an exemplary flowchart of an automatic calibration method for metrological data according to some embodiments of this application; Figure 2 This is a flowchart illustrating the process of determining the nonlinear compensation coefficient and hysteresis correction coefficient of a target smart sensor according to some embodiments of this application. Figure 3 This is a schematic diagram of the structure of an automatic calibration system for measurement data according to some embodiments of this application; Figure 4 This is a schematic diagram of the structure of a computer device for implementing an automatic calibration method for metrological data according to some embodiments of this application. Detailed Implementation

[0018] To better understand the technical solution of this application, the technical solution of this application will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0019] refer to Figure 1 The figure is an exemplary flowchart of an automatic calibration method for metrological data according to some embodiments of this application. The automatic calibration method for metrological data mainly includes the following steps: In step 101, during the normal operation of the metering equipment, the measured output value of the target intelligent sensor in the metering equipment for the measured physical quantity is acquired.

[0020] In this application, a calibration module is pre-integrated into the metrology equipment, and an externally traceable standard is connected. Specifically, this can be achieved by embedding a target intelligent sensor in the key measurement link of the metrology equipment to record the real-time response of the measured physical quantity. The sampling frequency of the intelligent sensor is set, for example, the sampling frequency can be set to 1 to 10 times per second to meet the dynamic monitoring requirements. This is only an example and is not intended to limit the specific scope of the invention. In addition, an external standard can be connected to the metrology equipment. The standard is used to provide a standard reference signal. This standard needs to be traced by a national metrology institution to ensure the accuracy of the measurement value. This will not be elaborated further here.

[0021] In practice, the measured output value of the physical quantity being measured by the target intelligent sensor in the metering equipment can be obtained in the following way: the electrical signal output of the physical quantity being measured is collected by the target intelligent sensor, the sampling frequency is, for example, 1 to 10 times per second, and the sensor is connected to the calibration module through a standard communication protocol (e.g., RS485) to collect and upload data periodically.

[0022] In addition, in specific implementation, the standard reference signal output by the external standard that has been traced by measurement value can be connected at the same time. That is, the external standard synchronously outputs a reference signal of the same type as the measured physical quantity. In specific implementation, it can be connected to the calibration module through a standard communication protocol and data can be uploaded at regular intervals. It should be noted that the measured output value in this application may specifically include voltage, current or digital value.

[0023] In some embodiments, the automatic calibration system for metrological data of this application is used in scenarios where multiple metrological devices are installed. The metrological devices monitor physical quantity data, and then transmit the collected data to a data processor. The data processor analyzes and processes the metrological data, performs automatic calibration, and finally transmits the calibration results to a data storage device for storage. This is only an example and is not intended to limit the specific scope of the invention.

[0024] In step 102, the standard reference value and standard threshold value pre-stored in the calibration module are read, and the standardized output value is determined based on the measured output value and the standard reference value.

[0025] Among them, the standard reference value is a real-time standard physical quantity value provided by the current external standard instrument through measurement traceability. It will be pre-stored in the storage unit of the calibration module to form a sequence. It also needs to determine the relevant threshold for the measurement range of the target smart sensor based on statistical analysis of historical measurement data, and compare it with the measured output value to judge the deviation.

[0026] The standard reference signal is the signal output by the external standard instrument with traceable measurement value when the measuring equipment is running normally. It corresponds to the measured output value of the physical quantity being measured by the target intelligent sensor. It is used to align the measured output value with the measured output value in time, extract the correlation mapping features, and assist in obtaining the standardized output value through fitting.

[0027] In some embodiments, reading the standard reference values ​​and standard thresholds pre-stored in the calibration module can be done in the following manner: obtaining a sequence of standard reference values ​​traced by measurement from the storage unit of the calibration module; determining the standard threshold based on statistical analysis of historical measurement data, and ensuring that the standard threshold is adapted to the measurement range of the target smart sensor, wherein the standard reference value is a real-time standard physical quantity value provided by the current external standard.

[0028] In addition, in some embodiments, determining the standardized output value based on the measured output value and the standard reference value can be achieved by the following steps: Multiple sets of measured output values ​​and multiple sets of standard reference values ​​are obtained from multiple sampling windows. The measured output values ​​and the standard reference values ​​are time-aligned, and the correlation mapping features between the measured output values ​​and the standard reference values ​​within the same sampling window are extracted. The dynamic correlation coefficient between the measured output value and the standard reference value is determined based on multiple sets of measured output values ​​and multiple sets of standard reference values. If the scatter plot of the measured output value and the standard reference value shows a linear trend, it is determined to be a linear deviation, and the Pearson correlation coefficient with a sliding window is used. If the scatter plot shows a monotonically non-linear trend, it is determined to be a non-linear deviation, and the Spearman rank correlation coefficient with a sliding window is used. The dynamic correlation between the measured output value and the standard reference value over time is analyzed, and the analysis result is used as the dynamic correlation coefficient. The standardized output value is obtained by weighting and summing all the correlation mapping features using all the dynamic correlation coefficients. Specifically, the standardized output value is obtained by weighting and summing the correlation mapping features using the correlation coefficient / residual standard deviation.

[0029] It should be noted that the correlation mapping feature in this application is a feature that measures the temporal synchronicity between the measured output value and the standard reference value; the dynamic correlation coefficient in this application is a statistic used to measure the strength of the local linear correlation between two sequences as they change over time within a sliding window.

[0030] In specific implementation, firstly, multiple sets of measured output values ​​and standard reference values ​​are obtained from multiple sampling windows. The standard reference signal is a physical quantity signal corresponding to the standard reference value, which is output in real time by an external standard instrument with traceable measurement values. The measured output values ​​and the standard reference values ​​are time-aligned. Extracting the correlation mapping features between the measured output values ​​and the standard reference signal within the same sampling window can be achieved in the following way: The measured output value and the standard reference signal are time aligned, that is, the two are resampled and interpolated according to the same sampling period to construct a data matrix under the same timestamp. Then, based on each time window, the mapping feature pairs between the measured output value and the standard reference signal are extracted. For example, the statistical features of normalized difference, moving average difference, and proportion can be used to form a feature vector, and this feature vector is used as the correlation mapping feature. Secondly, the dynamic correlation coefficient between the measured output value and the standard reference value, based on multiple sets of measured output values ​​and multiple sets of standard reference values, can be determined in the following way: The Pearson correlation coefficient analysis using a sliding window is used to analyze the dynamic correlation between measured output values ​​and standard reference values ​​over time. The analysis results are used as dynamic correlation coefficients. Specifically, data is extracted using a sliding window of fixed time length. For each window, the measured output value of the target intelligent sensor and the standard reference value for traceability are compared to the measured value to measure the degree of linear correlation between the two. As the window slides along the time axis, a series of correlation coefficients are obtained, which are the dynamic correlation coefficients, reflecting the correlation between the two over time. Then, the standardized output value can be obtained by weighting and summing all the correlation mapping features using all the dynamic correlation coefficients. This can be achieved by using the dynamic correlation coefficients as weights to perform weighted summation on the correlation mapping features, thus obtaining the standardized output value that reflects the bias correction.

[0031] In step 103, the standardized output value is compared with the standard threshold. When the standardized output value exceeds the standard threshold of the target smart sensor, the measurement device is controlled to switch to calibration mode. An external standard provides positive and negative standard input values ​​of known quantities, and the first and second response signals corresponding to the target smart sensor are collected simultaneously.

[0032] The calibration mode refers to the special calibration state that the metrology equipment switches into when the standardized output value of the target smart sensor exceeds a standard threshold. In some embodiments, the standardized output value is compared with the standard threshold. When the standardized output value exceeds the standard threshold of the target smart sensor, the metrology equipment is controlled to switch to calibration mode. An external standard provides positive and negative standard input values ​​of known quantities. The simultaneous acquisition of the first and second response signals corresponding to the target smart sensor can be achieved through the following steps: Compare the deviation between the measured output value and the standard reference value; When the deviation exceeds a preset threshold, the measuring device is triggered to enter calibration mode. The preset threshold (standard threshold) is calculated as: (total error of the sensor being calibrated, specified as either full-range error FS or indication error RV) × 0.8. For example, a 0.1-grade (full-range error FS) temperature sensor (range 0-100℃) has a permissible error of ±0.1℃, and the standard threshold is set to ±0.08℃; a 1.0-grade (indication error RV) pressure sensor has a permissible error of ±1.0% of the current indication, and the standard threshold is set to ±0.8% of the current indication. If the sensor's usage time exceeds 50% of the calibration cycle, and the environmental stress (temperature and humidity fluctuations > ±5℃, vibration frequency > 50Hz) or usage frequency (daily sampling times > 1000 times) exceeds the preset value, the standard threshold is lowered to permissible error × 0.6. A positive and negative standard input are then sequentially applied to the target smart sensor via an external standard. The first response signal is acquired when a positive standard input is applied, and the second response signal is acquired when a negative standard input is applied.

[0033] It should be noted that the degree of deviation in this application refers to the percentage difference between the measured output value and the standard reference value; while the positive standard input quantity in this application refers to the sequence of values ​​from small to large, and the negative standard input quantity refers to the sequence of values ​​from large to small.

[0034] In specific implementation, firstly, comparing the deviation between the measured output value and the standard reference value can be achieved by calculating the absolute difference between the measured output value and the standard reference value and dividing it by the standard reference value to obtain the deviation percentage. Secondly, when the deviation exceeds a preset threshold, the measuring device is triggered to enter the calibration mode, and positive and negative standard input quantities are sequentially applied to the target smart sensor through an external standard. This can be achieved by switching modes if the deviation percentage exceeds the threshold, for example, 0.5%, and controlling the external standard to apply a positive and negative sequence covering the range. The positive sequence covering the range is, for example, 0%, 25%, 50%, 75%, and 100% of the range, and the negative sequence covering the range is, for example, 100%, 75%, 50%, 25%, and 0% of the range. Then, acquiring the first response signal when applying the positive standard input quantity and the second response signal when applying the negative standard input quantity can be achieved by synchronously acquiring the sensor's response signal to each input point to ensure data timing consistency.

[0035] In step 104, based on the first response signal, the second response signal, and the corresponding positive and negative standard input quantities, the nonlinear compensation coefficient and hysteresis correction coefficient of the target smart sensor are calculated with reference to the standard reference value.

[0036] Among them, the nonlinear compensation coefficient refers to the coefficient calculated based on the deviation characteristics of the positive standard input quantity and the corresponding first response signal, which is used to compensate for the nonlinear deviation of the sensor, while the hysteresis correction coefficient is calculated based on the deviation characteristics of the positive / reverse standard input quantity and the corresponding response signal, which is used to correct the hysteresis deviation of the sensor.

[0037] In some embodiments, reference Figure 2 As shown in the figure, this is a flowchart illustrating the calculation of nonlinear compensation coefficients and hysteresis correction coefficients in some embodiments of this application. In this embodiment, the determination of the nonlinear compensation coefficients and hysteresis correction coefficients of the target smart sensor can be achieved through the following steps: In step 1031, based on the positive and negative standard input values ​​at multiple different range points, multiple sets of first and second response signals are simultaneously acquired. Specifically, for example, multiple sets of first and second response signals are acquired based on 7 evenly distributed range points (0%, 16.7%, 33.3%, 50%, 66.7%, 83.3%, 100%). In step 1032, the deviation distribution characteristics between the first response signal and the positive standard input, and the deviation distribution characteristics between the second response signal and the negative standard input are extracted. Specifically, for example, an external standard provides multiple positive and negative standard inputs at different range points, and the first and second response signals of the target smart sensor are collected simultaneously; the difference between the first response signal and the corresponding positive standard input, and the difference between the second response signal and the corresponding negative standard input are calculated respectively; statistical features such as mean and standard deviation are extracted from the above two types of differences, which are the deviation distribution characteristics. In step 1033, the nonlinear response intensity and hysteresis response intensity are determined based on the deviation distribution characteristics. The nonlinear response intensity is determined based on the deviation distribution characteristics (e.g., mean deviation, standard deviation) between the first response signal and the positive standard input. This characteristic is quantified; for example, the larger the deviation and the more discrete the distribution, the higher the intensity. The final quantized result is the nonlinear response intensity. The hysteresis response intensity is determined based on the deviation distribution characteristics between the second response signal and the reverse standard input. This characteristic is quantified by combining the difference between the positive and reverse deviations; for example, the larger the difference between the positive and reverse deviations, the higher the intensity. The final quantized result is the hysteresis response intensity. In step 1034, an initial nonlinear compensation factor is set based on the nonlinear response intensity: the nonlinear response intensity needs to be quantized based on the nonlinear model corresponding to the sensor model (dimensionless, value range 0-1, for example, platinum resistance sensor is quantized based on the resistance-temperature relationship model), and the initial nonlinear compensation factor = nonlinear response intensity × sensor full scale value / 10000 (for example, for a temperature sensor with a range of 0-100℃, when the nonlinear response intensity is 0.8, the initial nonlinear compensation factor = 0.8 × 100 / 10000 = 0.008); an initial hysteresis correction factor is set based on the hysteresis response intensity: the hysteresis response intensity needs to be quantized based on the hysteresis characteristic model corresponding to the sensor model, and the initial hysteresis correction factor = hysteresis response intensity × sensor full scale value / 10000. This is only an example and is not intended to limit the specific scope of the present invention. In step 1035, the first response signal is compensated using the initial nonlinear compensation factor, and the error between the compensated output value and the corresponding positive standard input is calculated. The step size is dynamically adjusted each time according to the current error gradient, where the error gradient refers to the rate of change of the global error (or consistency error) after two adjacent adjustments. When the error gradient is large, the step size is 1% of the initial factor; when the error gradient is small, the step size is reduced to 0.1% of the initial factor. The initial nonlinear compensation factor is repeatedly adjusted until the global error is reduced to the minimum (and ≤ 1 / 2 of the allowable error of the calibrated sensor), thereby obtaining the final nonlinear compensation coefficient. The hysteresis deviation between the first and second response signals is corrected using the initial hysteresis correction factor, and the consistency error of the corrected positive and negative response signals is calculated. The step size is dynamically adjusted according to the current error gradient (when the error gradient is large, the step size is 1% of the initial factor; when the error gradient is small, the step size is reduced to 0.1% of the initial factor). The initial hysteresis correction factor is repeatedly adjusted until the consistency error is reduced to the minimum (and ≤ 1 / 2 of the allowable error of the calibrated sensor), thereby obtaining the final hysteresis correction coefficient.

[0038] In specific implementation, for example, to calculate the nonlinear compensation coefficient, the first response signals of the seven uniformly distributed range points (0%, 16.7%, 33.3%, 50%, 66.7%, 83.3%, 100%) can be compensated one by one using the initial nonlinear compensation factor. The absolute error between the compensated output value and the corresponding positive standard input value at each range point is calculated, and the arithmetic mean of the errors of all range points is taken as the global error. Then, the step size is dynamically adjusted according to the current error gradient (when the error gradient is large, the step size is 1% of the initial factor; when the error gradient is small, the step size is reduced to 0.1% of the initial factor). The initial nonlinear compensation factor is repeatedly adjusted until the global error is reduced to the minimum, thereby obtaining the final nonlinear compensation coefficient.

[0039] In addition, the hysteresis correction coefficient can be calculated by successively correcting the difference (i.e., hysteresis deviation) between the first response signal and the second response signal at each of the seven uniformly distributed range points (0%, 16.7%, 33.3%, 50%, 66.7%, 83.3%, 100%) using the initial hysteresis correction factor. The absolute difference between the first response signal and the second response signal after correction is calculated for each range point, and the arithmetic mean of this absolute difference for all range points is taken as the consistency error. Then, the step size is dynamically adjusted according to the current error gradient (when the error gradient is large, the step size is 1% of the initial factor; when the error gradient is small, the step size is reduced to 0.1% of the initial factor). The initial hysteresis correction factor is repeatedly adjusted until the consistency error is reduced to the minimum, thereby obtaining the final hysteresis correction coefficient.

[0040] It should be noted that the deviation distribution characteristics in this application are characteristics that measure the degree of deviation between the response signal and the standard input; the nonlinear response intensity refers to a quantitative index that reflects the deviation of the sensor output from the ideal linearity; and the hysteresis response intensity refers to a quantitative index that reflects the difference between the positive and negative responses.

[0041] Furthermore, it should be noted that this application introduces multi-range point acquisition to quantitatively model sensor deviations, realizing the conversion from quantitative data to compensation coefficients, effectively improving the inclusiveness and flexibility of the error range. Secondly, by integrating nonlinear and hysteresis information through an error minimization mechanism, independent coefficients are obtained, enhancing the ability to express the coupling relationship of multi-source errors. Finally, compensation coefficients are output based on real-time standard values, providing a path for dynamic, flexible, and interpretable characterization of errors, thereby improving the accuracy of identifying deviation risks in metrological calibration. This scheme not only improves the accuracy of calibration but also enhances the system's adaptability to uncertainties and error differences.

[0042] In step 105, the output correction parameters of the target smart sensor are updated according to the nonlinear compensation coefficient and the hysteresis correction coefficient.

[0043] In some embodiments, updating the output correction parameters of the target smart sensor based on the nonlinear compensation coefficient and the hysteresis correction coefficient can be achieved by the following steps: The nonlinear compensation coefficient is applied to the output function of the target intelligent sensor to compensate for the nonlinear deviation. The hysteresis correction coefficient is applied to the output function of the target smart sensor to correct the hysteresis deviation; Based on the compensated and corrected output function, the output correction parameter is adjusted by combining multiple sets of historical calibration data. The retrieved historical calibration data must meet the requirements of the same standard input and the same environmental conditions (environmental conditions include temperature 25±2℃, humidity 40%~60% RH, and no electromagnetic interference). The data content includes the historical standard input, the corresponding sensor response signal, and the output correction parameter at that time. The current sensor response to the same standard input (the current environmental conditions must be consistent with the historical data environmental conditions) is substituted into the corrected output function and compared with the corrected output under the same historical conditions. The deviation is analyzed and the output correction parameter is adjusted to adapt to the current sensor response characteristics.

[0044] It should be noted that the output function in this application refers to the transformation relationship between the sensor's input physical quantity and its output value; the output correction parameter in this application refers to the weights or biases used to adjust the output function.

[0045] In practical implementation, firstly, the nonlinear compensation coefficient is applied to the output function of the target intelligent sensor. Compensation for nonlinear deviation can be achieved as follows: the target intelligent sensor exhibits nonlinear deviation characteristics within its measurement range. Specifically, under different input values ​​of the measured physical quantity, the deviation between the sensor's original output value and the true value of the measured physical quantity shows a non-constant trend (i.e., as the value of the measured physical quantity increases, the deviation amplitude nonlinearly increases or decreases). The nonlinear compensation coefficient obtained through iterative adjustment by minimizing the error is the core parameter used to correct this nonlinear deviation. In practical implementation, this nonlinear compensation coefficient needs to be integrated into the original output function of the target intelligent sensor. For example, it can be integrated into the original output function by adding a compensation term or correcting the coefficient parameters in the function. It should be noted that this original output function is used to construct the mapping relationship between the sensor's measured electrical signal and the output physical quantity (e.g., temperature, pressure). The output function, after being corrected by the nonlinear compensation coefficient, can correct the deviation of the original output value of the sensor across the entire range, significantly reducing its nonlinear deviation amplitude. This ensures that the deviation between the corrected output value and the true value of the measured physical quantity is controlled within the allowable range, ultimately achieving accurate compensation for the nonlinear deviation of the target intelligent sensor.

[0046] Furthermore, adjusting the output correction parameters based on the compensated and corrected output function, combined with multiple sets of historical calibration data, to adapt it to the current sensor's response characteristics can be achieved in the following way: First, using the corrected output function with completed nonlinear compensation as a basis, retrieve multiple sets of historical calibration data (including historical standard input quantities, corresponding sensor response signals, and the output correction parameters at that time); then, substitute the current sensor's response to the same standard input into the corrected output function, compare it with the corrected output under the same historical conditions, and analyze the deviation; then, iteratively adjust the output correction parameters based on the deviation (e.g., fine-tune the compensation correlation coefficient and the coefficient of the correction function term); finally, ensure that the adjusted parameters allow the current sensor output to meet the deviation from the standard value, thus adapting to its current response characteristics.

[0047] It should be noted that the proposed solution updates the output correction parameters based on the compensation coefficient, which effectively avoids the excessive contribution of error in the sensor, improves the generalization ability and stability of calibration, and compared with the fixed or unconstrained adjustment of error weights in the prior art, this solution dynamically suppresses abnormal deviations and reduces the impact of noise on the results, thereby improving the overall calibration accuracy and robustness and meeting the diverse and nonlinear requirements of complex metrology scenarios.

[0048] In addition, this application also provides an automatic calibration system for metrological data, with reference to Figure 3The figure is a schematic diagram of the structure of an automatic calibration system for metrological data according to some embodiments of this application. The automatic calibration system 400 for metrological data includes: an acquisition module 401, a processing module 402, and an execution module 403, which are described below: The acquisition module 401 in this application is mainly used to acquire the measured output value of the target intelligent sensor of the metering equipment for the measured physical quantity during the normal operation of the metering equipment. Processing module 402 in this application is mainly used to read the standard reference value and standard threshold value pre-stored in the calibration module, and determine the standardized output value based on the measured output value and the standard reference value; It should be noted that the processing module 402 described in this application is also used to compare the standardized output value with the standard threshold. When the standardized output value exceeds the standard threshold of the target smart sensor, the measurement device is controlled to switch to calibration mode, and a positive standard input and a negative standard input of known values ​​are provided by an external standard, and the first response signal and the second response signal corresponding to the target smart sensor are collected simultaneously. In addition, the processing module 402 is also used to calculate the nonlinear compensation coefficient and hysteresis correction coefficient of the target smart sensor based on the first response signal, the second response signal and the corresponding positive and negative standard input quantities, with reference to the standard reference value. The execution module 403 in this application is mainly used to update the output correction parameters of the target smart sensor according to the nonlinear compensation coefficient and the hysteresis correction coefficient.

[0049] In addition, this application also provides a computer device, the computer device including a memory and a processor, the memory storing code, and the processor being configured to acquire the code and execute the above-described automatic calibration method for metrological data.

[0050] In some embodiments, reference Figure 4 The figure is a schematic diagram of the structure of a computer device implementing an automatic calibration method for metrological data according to some embodiments of this application. The automatic calibration method for metrological data in the above embodiments can be achieved through... Figure 4 The computer device 500 shown is used to implement this, and the computer device 500 includes at least one processor 501, a communication bus 502, a memory 503, and at least one communication interface 504.

[0051] Processor 501 can be a general-purpose central processing unit (CPU) or an application-specific integrated circuit (ASIC).

[0052] The communication bus 502 can be used to transmit information between the aforementioned components.

[0053] Memory 503 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CDROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital versatile optical discs, Blu-ray discs, etc.), magnetic disks or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. Memory 503 may exist independently and be connected to processor 501 via communication bus 502. Memory 503 may also be integrated with processor 501.

[0054] The memory 503 stores program code for executing the solution of this application, and its execution is controlled by the processor 501. The processor 501 executes the program code stored in the memory 503. The program code may include one or more software modules. In the above embodiments, the automatic calibration method for metrological data can be implemented by the processor 501 and one or more software modules in the program code in the memory 503.

[0055] Communication interface 504 uses any transceiver-like device to communicate with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area networks (WLAN), etc.

[0056] In a specific implementation, as one example, a computer device may include multiple processors, each of which may be a single-core (single CPU) processor or a multi-core (multi CPU) processor. Here, a processor may refer to one or more devices, circuits, and / or processing cores used to process data (e.g., computer program instructions).

[0057] The aforementioned computer device can be a general-purpose computer device or a special-purpose computer device. In specific implementations, the computer device can be a desktop computer, a portable computer, a network server, a handheld digital assistant (PDA), a mobile phone, a tablet computer, a wireless terminal device, a communication device, or an embedded device. This application does not limit the type of computer device.

[0058] In addition, this application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described automatic calibration method for metrological data.

[0059] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0060] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method of automatic calibration of metrology data, characterized by, The application relates to a method for calibrating a target intelligent sensor in a metering device. During normal operation of the metering device, an actual measured output value of a target intelligent sensor in the metering device for a measured physical quantity is acquired; a standard reference value and a standard threshold value pre-stored in a calibration module are read, and a standardized output value is determined according to the actual measured output value and the standard reference value; the standardized output value is compared with the standard threshold value, when the standardized output value exceeds the standard threshold value of the target intelligent sensor, the metering device is controlled to switch to a calibration mode, a known positive standard input quantity and a known reverse standard input quantity are provided by an external standard device, and a first response signal and a second response signal corresponding to the target intelligent sensor are synchronously acquired; a nonlinear compensation coefficient and a hysteresis correction coefficient of the target intelligent sensor are calculated according to the first response signal, the second response signal and the corresponding positive standard input quantity and reverse standard input quantity, and the standard reference value is referred to; the output correction parameter of the target intelligent sensor is updated according to the nonlinear compensation coefficient and the hysteresis correction coefficient.

2. The method of claim 1, wherein, The determination of the standardized output value according to the actual measured output value and the standard reference value specifically includes: a plurality of groups of actual measured output values and standard reference values are acquired from a plurality of sampling windows, the actual measured output values and the standard reference values are time-aligned, and the correlation mapping features of the actual measured output values and the standard reference signals in the same sampling window are extracted; dynamic correlation coefficients of the actual measured output values and the standard reference values are determined based on the plurality of groups of actual measured output values and the plurality of groups of standard reference values; all the correlation mapping features are weighted and summed to obtain the standardized output value.

3. The method of claim 1, wherein, The comparison of the standardized output value with the standard threshold value, when the standardized output value exceeds the standard threshold value of the target intelligent sensor, the metering device is controlled to switch to a calibration mode, a known positive standard input quantity and a known reverse standard input quantity are provided by an external standard device, and a first response signal and a second response signal corresponding to the target intelligent sensor are synchronously acquired specifically includes: the deviation degree of the actual measured output value and the standard reference value is compared; when the deviation exceeds a preset threshold value, the metering device is triggered to enter the calibration mode, and the external standard device is used to sequentially apply the positive standard input quantity and the reverse standard input quantity to the target intelligent sensor; the first response signal is acquired when the positive standard input quantity is applied, and the second response signal is acquired when the reverse standard input quantity is applied.

4. The method of claim 1, wherein, The calculation of the nonlinear compensation coefficient and the hysteresis correction coefficient of the target intelligent sensor based on the first response signal, the second response signal and the corresponding positive standard input quantity and reverse standard input quantity, and the standard reference value is referred to specifically includes: a plurality of groups of first response signals and second response signals are synchronously acquired based on the positive standard input quantity and the reverse standard input quantity of a plurality of different range points; deviation distribution features between the first response signal and the positive standard input quantity and deviation distribution features between the second response signal and the reverse standard input quantity are extracted; the nonlinear response intensity and the hysteresis response intensity are determined based on the deviation distribution features; an initial nonlinear compensation factor is set according to the nonlinear response intensity, and an initial hysteresis correction factor is set according to the hysteresis response intensity; The nonlinear compensation coefficient is obtained by calculating the error between the initial nonlinear compensation factor and the real-time standard value provided by the current external standard device, and repeatedly adjusting the initial nonlinear compensation factor to minimize the error; and the hysteresis correction coefficient is obtained by calculating the error between the initial hysteresis correction factor and the real-time standard value provided by the current external standard device, and repeatedly adjusting the initial hysteresis correction factor to minimize the error.

5. The method of claim 1, wherein, The updating of the output correction parameter of the target intelligent sensor according to the nonlinear compensation coefficient and the hysteresis correction coefficient specifically comprises: applying the nonlinear compensation coefficient to the output function of the target intelligent sensor to compensate for the nonlinear deviation; applying the hysteresis correction coefficient to the output function of the target intelligent sensor to correct the hysteresis deviation; based on the compensated and corrected output function, adjusting the value of the output correction parameter to adapt to the response characteristics of the current sensor in combination with a plurality of sets of historical calibration data.

6. The method of claim 1, wherein, The reading of the standard reference value and the standard threshold value pre-stored in the calibration module specifically comprises: obtaining the standard reference value sequence traceable in quantity from the storage unit of the calibration module; determining the standard threshold value based on statistical analysis of historical metrological data, and ensuring that the standard threshold value adapts to the range of the target intelligent sensor, wherein the standard reference value is a real-time standard physical value provided by the current external standard device.

7. The method of claim 1, wherein, The external standard device is specifically a precision calibration instrument or a reference measurement device capable of providing a standard physical value traceable in quantity.

8. A metrology data auto-calibration system, characterized by, comprises: an acquisition module, configured to acquire a measured output value of a target intelligent sensor for a measured physical quantity in a metrological device during normal operation of the metrological device; a processing module, configured to read a standard reference value and a standard threshold value pre-stored in a calibration module, and determine a standardized output value according to the measured output value and the standard reference value; the processing module is further configured to compare the standardized output value with the standard threshold value, and control the metrological device to switch to a calibration mode when the standardized output value exceeds the standard threshold value of the target intelligent sensor, so that the external standard device provides a forward standard input quantity and a reverse standard input quantity of a known quantity, and synchronously acquires a first response signal and a second response signal corresponding to the target intelligent sensor; the processing module is further configured to calculate a nonlinear compensation coefficient and a hysteresis correction coefficient of the target intelligent sensor with reference to the standard reference value based on the first response signal, the second response signal, and the corresponding forward standard input quantity and reverse standard input quantity; an execution module, configured to update an output correction parameter of the target intelligent sensor according to the nonlinear compensation coefficient and the hysteresis correction coefficient. 9.A computer device, comprising a memory and a processor, wherein the memory stores code, and the code comprises the following steps: The processor is configured to acquire the code and execute the metrological data automatic calibration method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, the computer program comprising instructions that, when executed by a computer, cause the computer to perform the method of any one of claims 1 to 9. The computer program is executed by the processor to implement the metrological data automatic calibration method according to any one of claims 1 to 7.

Images