Dynamic calibration error compensation method for electrical measuring instrument

Abstract

The invention relates to the technical field of electrical measurement, and discloses an electrical measurement instrument dynamic calibration error compensation method, which comprises the following steps: S1, multi-source data real-time acquisition: integrating a temperature sensor, a humidity sensor and an electromagnetic interference detection module in a measurement instrument, configuring voltage, current, frequency and other measurement channels, and carrying out multi-source data real-time acquisition at the frequency of not less than 100HZ; environmental parameters (temperature T, humidity H and electromagnetic interference intensity E) and measurement data (model characteristic parameters X1, X2,..., Xn such as voltage, current and frequency) are synchronously acquired. According to the electrical measuring instrument dynamic calibration error compensation method, through multi-source data real-time acquisition and composite error model construction, in combination with multiple linear regression and an improved BP neural network, environment and measurement signal changes are comprehensively captured, an error compensation value is accurately calculated, a measurement error can be effectively controlled within an extremely small range, and the error compensation accuracy is improved. Compared with a traditional method, the measurement precision is greatly improved.

Description

Technical Field

[0001] The present invention relates to the technical field of electrical measurement, and in particular to a method for compensating dynamic calibration errors of electrical measuring instruments. Background Art

[0002] In the field of electrical measurement, the accuracy of measuring instruments is of vital importance. Existing electrical measuring instruments are affected by various factors during long-term use, such as changes in ambient temperature and humidity, aging of components in the measurement circuit, and electromagnetic interference during the measurement process. These factors can cause measurement errors to gradually increase, seriously affecting the reliability of the measurement results.

[0003] Traditional calibration methods typically use high-precision standard sources to statically calibrate measuring instruments in a specific standard environment. However, the actual working environment is often complex and changeable. This static calibration method cannot adapt to dynamic changes in the environment and measurement conditions in real time, and it is difficult to effectively compensate for dynamic errors generated during the measurement process.

[0004] For example, in industrial sites, the ambient temperature may fluctuate significantly in a short period of time. The inductance of an ordinary ferrite inductor may change by about 2% due to temperature changes. If static calibration is relied upon alone, the measurement error cannot be effectively controlled. At the same time, there is a large amount of electromagnetic interference in industrial sites, such as 50Hz power frequency interference and its harmonics, which will cause fluctuations in measurement results. Existing technologies are difficult to guarantee measurement accuracy in such complex interference environments. In addition, the calibration process of existing measuring instruments is often more complicated, requiring professional operation, and the calibration cycle is long, affecting production efficiency. Moreover, some calibration methods have high requirements for calibration equipment, which increases cost investment.

[0005] Therefore, a dynamic calibration error compensation method for electrical measuring instruments is proposed to solve the above-mentioned problems. Summary of the Invention

[0006] (1) Technical problems solved

[0007] In response to the shortcomings of the existing technology, the present invention provides a dynamic calibration error compensation method for electrical measuring instruments, which has the advantages of being able to monitor changes in the environment and measurement conditions in real time and dynamically compensate for measurement errors. It solves the problem that traditional calibration methods usually use high-precision standard sources to statically calibrate measuring instruments under a specific standard environment; however, the actual working environment is often complex and changeable, and this static calibration method cannot adapt to the dynamic changes in the environment and measurement conditions in real time, making it difficult to effectively compensate for the dynamic errors generated during the measurement process.

[0008] (2) Technical solution

[0009] In order to achieve the above-mentioned purpose of being able to monitor changes in the environment and measurement conditions in real time and dynamically compensate for measurement errors, the present invention provides the following technical solution: a method for dynamic calibration error compensation of an electrical measuring instrument, comprising the following steps:

[0010] S1: Real-time acquisition of multi-source data: Integrate temperature, humidity sensors and electromagnetic interference detection modules in the measuring instrument, configure voltage, current, frequency and other measurement channels, and synchronously collect environmental parameters (temperature T, humidity H, electromagnetic interference intensity E) and measurement data (voltage, current, frequency and other model characteristic parameters X1, X2, ..., X) at a frequency of not less than 100HZ. n );

[0011] S2: Data preprocessing: Use IIR / FIR digital filter combination to process the measurement signal, improve the signal-to-noise ratio to more than 100dB for 50Hz power frequency interference and harmonic notch; smooth the temperature and humidity data and store the processed data;

[0012] S3: Construct a composite error model: Use the multivariate linear regression algorithm to build a linear relationship based on the experimental data to form the basic model ΔY linear =f linear (T,H,E,X1,X2,…,X n ); Use machine learning algorithms to mine nonlinear relationships and build a prediction model ΔY nonlinear =f nonlinear (T,H,E,X1,X2,…,X n ), the two are added together to get the composite model ΔY=ΔY linear +ΔY nonlinear ;

[0013] S4: Dynamic error compensation calculation and implementation: Substitute the processed parameters into the composite model, calculate the compensation value ΔY, and correct the original result Y through the hardware feedback circuit. raw , get Y compensated =Y raw -ΔY;

[0014] S5: Intelligent optimization of calibration parameters: Uses machine learning algorithms to regularly analyze data and automatically optimize model coefficients when compensation errors exceed 0.5% or environmental changes trigger conditions;

[0015] S6: Abnormal alarm and processing: If ΔY exceeds the range during compensation, the instrument will sound an alarm and prompt you to check the instrument and environment.

[0016] Preferably, the temperature sensor has an accuracy of ±0.1°C, the humidity sensor has an accuracy of ±2%RH, and the electromagnetic interference detection module has an accuracy of ±1dB.

[0017] Preferably, in the multi-source data real-time acquisition step, when the environmental parameters change by more than ±2°C in temperature, ±5%RH in humidity, or ±5dB in electromagnetic interference intensity, high-speed data acquisition is automatically triggered, and the acquisition frequency is increased to 500Hz.

[0018] Preferably, in the IIR / FIR digital filter combination, the order of the IIR filter is 8, the number of taps of the FIR filter is 32, and the notch frequency range is 49 Hz-51 Hz.

[0019] Preferably, in the step of constructing the composite error model, the experimental data is updated after each 1000 dynamic error compensation calculations are performed, and the basic error compensation model and the error prediction model are reconstructed based on the updated data.

[0020] Preferably, the machine learning algorithm adopts an improved BP neural network, which includes three hidden layers, the number of neurons is 16-32-16 respectively, the activation function is a ReLU-Sigmoid hybrid function, the Adam optimization algorithm is used during training, and the learning rate is dynamically adjusted in the range of 0.001-0.01.

[0021] Preferably, the hardware feedback circuit includes a DSP core processing unit, a 16-bit DAC conversion chip and a differential amplifier circuit. The compensation signal is connected to the analog front end after being isolated by a photoelectric coupling, and the compensation delay time is ≤20ms.

[0022] Preferably, in the multi-source data real-time acquisition step, the measurement data and environmental parameters are transmitted via the SPI protocol with a transmission rate of 5 Mbps, and the data frame includes a 16-bit check code for data integrity verification.

[0023] Preferably, the DSP core processing unit, 16-bit DAC conversion chip and differential amplifier circuit in the hardware feedback circuit are synchronized using a unified 100MHz clock source, with a clock jitter of ≤100ps, to ensure phase consistency between the compensation signal and the measurement signal.

[0024] (3) Beneficial effects

[0025] Compared with the prior art, the present invention provides a method for compensating for dynamic calibration errors of electrical measuring instruments, which has the following beneficial effects:

[0026] 1. This dynamic calibration error compensation method for electrical measuring instruments uses real-time multi-source data acquisition and composite error model construction, combined with multiple linear regression and an improved BP neural network, to comprehensively capture changes in the environment and measurement signals, accurately calculate error compensation values, and effectively control measurement errors to an extremely small range. Compared with traditional methods, it significantly improves measurement accuracy. For example, when measuring 220V voltage, it can correct a 0.5V error to ensure the reliability of measurement results.

[0027] 2. This dynamic calibration error compensation method for electrical measuring instruments integrates multiple sensors to monitor environmental parameters such as temperature, humidity, and electromagnetic interference in real time. When environmental changes exceed the threshold, the acquisition frequency is automatically increased. Combined with targeted data preprocessing and intelligent optimization of calibration parameters, it can adapt to complex and changeable environments such as industrial workshops, effectively deal with the impact of factors such as temperature fluctuations and electromagnetic interference on measurement accuracy, and broaden the application scenarios of the instrument.

[0028] 3. This dynamic calibration error compensation method for electrical measuring instruments features highly automated data processing and calibration processes, reducing the need for professional personnel. It uses common sensors and mature hardware and algorithms to reduce reliance on high-precision calibration equipment, simplify the calibration process, and shorten the calibration cycle. Meanwhile, designs such as cyclic storage optimize resource utilization, reduce equipment use and maintenance costs, and improve production efficiency.

[0029] 4. This electrical measuring instrument's dynamic calibration error compensation method uses a machine learning algorithm to intelligently optimize calibration parameters, enabling the error compensation model to adapt to long-term changes in the environment and measurement conditions. The abnormal alarm and processing mechanism provides timely reminders when the error compensation value exceeds the range. The synchronous hardware circuit design and electromagnetic shielding protection structure ensure the accuracy of the compensation signal and the instrument's anti-interference ability, thereby improving system stability and reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 This is a flow chart of the method for dynamic calibration error compensation of an electrical measuring instrument according to the present invention;

[0031] Figure 2 This is the overall workflow diagram of the dynamic calibration error compensation method for electrical measuring instruments of the present invention. DETAILED DESCRIPTION

[0032] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0033] Example 1

[0034] See also Figure 1-2 , a method for compensating dynamic calibration errors of electrical measuring instruments, comprising the following steps:

[0035] S1: Real-time acquisition of multi-source data: Integrate temperature, humidity sensors and electromagnetic interference detection modules in the measuring instrument, configure voltage, current, frequency and other measurement channels, and synchronously collect environmental parameters (temperature T, humidity H, electromagnetic interference intensity E) and measurement data (voltage, current, frequency and other model characteristic parameters X1, X2, ..., X) at a frequency of not less than 100HZ. n );

[0036] S2: Data preprocessing: Use IIR / FIR digital filter combination to process the measurement signal, improve the signal-to-noise ratio to more than 100dB for 50Hz power frequency interference and harmonic notch; smooth the temperature and humidity data and store the processed data;

[0037] S3: Construct a composite error model: Use the multivariate linear regression algorithm to build a linear relationship based on the experimental data to form the basic model ΔY linear =f linear (T,H,E,X1,X2,…,X n ); Use machine learning algorithms to mine nonlinear relationships and build a prediction model ΔY nonlinear =f nonlinear (T,H,E,X1,X2,…,X n ), the two are added together to get the composite model ΔY=ΔY linear +ΔY nonlinear ;

[0038] S4: Dynamic error compensation calculation and implementation: Substitute the processed parameters into the composite model, calculate the compensation value ΔY, and correct the original result Y through the hardware feedback circuit. raw , get Y compensated =Y raw -ΔY;

[0039] S5: Intelligent optimization of calibration parameters: Uses machine learning algorithms to regularly analyze data and automatically optimize model coefficients when compensation errors exceed 0.5% or environmental changes trigger conditions;

[0040] S6: Abnormal alarm and processing: If ΔY exceeds the range during compensation, the instrument will sound an alarm and prompt you to check the instrument and environment.

[0041] This dynamic calibration error compensation method for electrical measuring instruments uses real-time data collection from multiple sources to comprehensively capture changes in the environment and measurement signals, laying the foundation for accurate compensation. Data preprocessing effectively filters out interference to ensure data quality. The composite error model combines linear and nonlinear relationships to accurately characterize error patterns. Dynamic compensation and hardware feedback circuits implement real-time corrections, significantly improving measurement accuracy. Intelligent optimization of calibration parameters enables the model to adapt to long-term environmental changes. The abnormal alarm mechanism enhances system reliability, and the coordinated operation of multiple links can control measurement errors to an extremely small range, significantly improving the measurement accuracy and reliability of instruments in complex environments, reducing calibration complexity and costs, and providing accurate and stable measurement data support for scenarios such as industrial production and scientific research testing.

[0042] Before executing the multi-source data real-time acquisition step S1, the measuring instrument performs an initialization operation: a self-test is performed on the integrated temperature sensor, humidity sensor, electromagnetic interference detection module, and measurement channels such as voltage, current, and frequency. If a sensor or channel fault is detected, a fault code is displayed on the display screen, and the fault information is recorded in the storage module of the measuring instrument. The correspondence between the fault code and the fault type is pre-stored in the EEPROM of the instrument.

[0043] The experimental data are derived from at least 1,000 sets of measurement experiments under different working conditions on the measuring instrument in different standard environmental chambers (temperature range -40°C-85°C, humidity range 10% RH-90% RH) using high-precision standard sources (voltage accuracy ±0.01%, current accuracy ±0.02%, frequency accuracy ±0.001Hz). Each set of experiments collects environmental parameters and measurement data at no less than 50 sample points.

[0044] The measuring instrument is provided with a system maintenance interface, through which stored processed data, model parameters, calibration records and other information can be exported; at the same time, the interface supports the import of new error compensation model algorithm programs for upgrading the error compensation function of the measuring instrument.

[0045] Example 2

[0046] 1. Multiple Linear Regression Algorithm

[0047] The multiple linear regression algorithm is used in the basic model stage of building the composite error model;

[0048] The principle is to assume that the measurement error is related to the environmental parameters (temperature T, humidity H, electromagnetic interference intensity E) and the characteristic parameters of the measurement signal (X1, X2, ..., X n ) has a linear relationship, and the expression is where ΔY l inear is the measurement error of the linear model prediction, β0,β1,…,βn+3 is the regression coefficient to be determined; environmental parameters, measurement signal characteristic parameters and corresponding actual measurement error data are obtained through a large number of experiments, and the regression coefficient is solved using the least squares method; the goal of the least squares method is to minimize the sum of squares of the prediction error, that is, S(β) = ∑ i-1 (ΔY j -(β0+β1T j +β2H j +β3E j +∑ i-1 β i+3 X ij )) 2 Minimize, where m is the number of samples of experimental data; by taking the partial derivative of S(β) with respect to each regression coefficient and setting it to zero, the optimal solution of the regression coefficient can be obtained, thereby determining the basic error compensation model.

[0049] 2. Improved BP Neural Network Algorithm

[0050] It is used to mine the nonlinear relationship between environmental parameters and measurement errors and build an error prediction model. The network consists of an input layer, three hidden layers (the number of neurons is 16-32-16 respectively) and an output layer. The input layer receives environmental parameters (temperature T, humidity H, electromagnetic interference intensity E) and measurement signal characteristic parameters (X1, X2, ..., X n ), the output layer outputs the predicted measurement error ΔY nonlinear The activation function uses a ReLU-Sigmoid hybrid function. The ReLU function (f(x) = max(0, x)) is used in the first half of the hidden layer to speed up training and alleviate the gradient disappearance problem; the Sigmoid function is used in the second half. Map the data to the (0,1) interval for easy processing by the output layer.

[0051] The training process uses the Adam optimization algorithm, which combines the advantages of the AdaGrad and RMSProp algorithms and adaptively adjusts the learning rate of each parameter. The dynamic adjustment range of the learning rate is 0.001-0.01. A larger learning rate is used to accelerate the convergence speed in the early stage of training, and the learning rate is gradually reduced as the training progresses to avoid missing the optimal solution. During training, the experimental data is input into the network, and the mean square error between the output and the actual measurement error is calculated. The error signal is propagated from the output layer to the input layer through the back propagation algorithm, and the connection weights and thresholds between the neurons in each layer are updated. It is continuously iterated until the mean square error reaches a satisfactory level, thereby determining the error prediction model.

[0052] 3. Least Squares Method (for intelligent optimization of calibration parameters)

[0053] In the calibration parameter intelligent optimization step, the coefficients of the composite error model are optimized and adjusted using the least squares method; the process is initiated when the compensation error exceeds a set threshold (such as 0.5%) or meets trigger conditions such as environmental changes.

[0054] The principle is to use the composite error model ΔY=ΔY based on the current measurement data and actual measurement error. linear +ΔY nonlinear The coefficients in are used as variables. By minimizing the sum of squares of prediction errors, the regression coefficients of the multivariate linear regression model and the connection weights and thresholds of the neural network are recalculated and updated, so that the composite error model can better adapt to changes in the environment and measurement conditions and improve the accuracy of error compensation.

[0055] 4. Digital filtering algorithm (IIR / FIR digital filter combination)

[0056] In the data preprocessing step, an IIR filter and an FIR filter are used in combination. The IIR filter is an 8th-order filter. By utilizing its recursive structure, feedback operation is performed on the input signal, which can achieve highly selective filtering characteristics at a lower order and effectively suppress 50Hz power frequency interference and its harmonics.

[0057] The FIR filter has 32 taps and, through non-recursive convolution operations, exhibits linear phase characteristics, ensuring that the phase relationship between the signal's frequency components remains unchanged. It further smoothes the signal and removes high-frequency noise and residual interference. The notch frequency range is set to 49Hz-51Hz, accurately filtering out 50Hz power frequency interference and its harmonics, increasing the signal-to-noise ratio to over 100dB and providing high-quality data for subsequent error compensation.

[0058] Example 3

[0059] Simulate test scenarios

[0060] 1. Implementation Environment Construction

[0061] An industrial high-precision electrical measuring instrument was selected as the implementation carrier. The instrument integrates a temperature sensor with an accuracy of ±0.1℃, a humidity sensor with an accuracy of ±2%RH, and an electromagnetic interference detection module with an accuracy of ±1dB. It is also equipped with a voltage measurement channel (range 0-1000V), a current measurement channel (range 0-50A), and a frequency measurement channel (range 10Hz-100kHz). The hardware feedback circuit uses a DSP core processing unit (model TMS320F28335), a 16-bit DAC conversion chip (model AD5660), and a differential amplifier circuit (model AD8138). The compensation signal is connected to the analog front end through an optocoupler isolation chip (model HCPL-7840) to ensure that the compensation delay time is ≤20ms.

[0062] 2. Real-time collection of multi-source data (S1)

[0063] After the instrument is put into use, it synchronously collects environmental parameters (temperature T, humidity H, electromagnetic interference intensity E) and measurement data (signal characteristic parameters such as voltage, current, and frequency) at a frequency of 100Hz; when the environmental parameters change by more than ±2°C for temperature, ±5%RH for humidity, and ±5dB for electromagnetic interference intensity, it automatically triggers high-speed data acquisition, and the acquisition frequency is increased to 500Hz; the measurement data and environmental parameters are transmitted via the SPI protocol with a transmission rate of 5Mbps, and the data frame contains a 16-bit check code for data integrity verification; for example, in a measurement scenario in an industrial workshop, the initial ambient temperature is 25°C, the humidity is 40%RH, and the electromagnetic interference intensity is 20dB. After the equipment starts running for a period of time, the temperature rises to 28°C, exceeding the temperature change threshold. At this time, the instrument automatically increases the acquisition frequency to 500Hz to obtain more accurate data in real time.

[0064] 3. Data Preprocessing (S2)

[0065] The collected data enters the data preprocessing stage, and the measurement signal is processed using a combination of an 8th-order IIR filter and a 32-tap FIR filter. The notch frequency range is set to 49Hz-51Hz, effectively filtering out 50Hz power frequency interference and its harmonics, and improving the signal-to-noise ratio to above 100dB. For temperature and humidity data, the sliding average filter method is used for smoothing to remove data fluctuation interference. The processed data is stored in the instrument's local storage module with a storage capacity of 1GB. A circular storage method is used. When the storage space is insufficient, the oldest stored data is automatically overwritten.

[0066] 4. Constructing a composite error model (S3)

[0067] 1. Construction of basic error compensation model: During the instrument development stage, 1,000 sets of measurement experiments under different working conditions were conducted on the measuring instrument in different standard environmental chambers (temperature range -40°C-85°C, humidity range 10%RH-90%RH) using high-precision standard sources (voltage accuracy ±0.01%, current accuracy ±0.02%, frequency accuracy ±0.001Hz). Each set of experiments collected no less than 50 sample points of environmental parameters and measurement data; based on these experimental data, a multivariate linear regression algorithm was used to solve the regression coefficient through the least squares method, and a linear relationship between the measurement error and the environmental parameters and the characteristic parameters of the measurement signal was established to form a basic error compensation model.

[0068] 2. Error prediction model construction: The experimental data is input into the improved BP neural network, which contains three hidden layers with 16-32-16 neurons respectively. The activation function adopts the ReLU-Sigmoid hybrid function. The Adam optimization algorithm is used for training, and the learning rate is dynamically adjusted in the range of 0.001-0.01. The back propagation algorithm is used to continuously send iterations for training, calculate the mean square error between the output and the actual measurement error, and update the connection weights and thresholds between neurons in each layer until the mean square error reaches a satisfactory level. The error prediction model ΔY is established. nonlinear .

[0069] 3. Composite error model generation: Add the basic error compensation model and the error prediction model to obtain the composite error model ΔY = ΔY linear +ΔY nonlinear During the operation of the instrument, the experimental data is updated after every 1,000 dynamic error compensation calculations, and the basic error compensation model and error prediction model are rebuilt based on the updated data to ensure the accuracy of the model.

[0070] 5. Dynamic Error Compensation Calculation and Implementation (S4)

[0071] Substitute the environmental parameters and measurement signal characteristic parameters after data preprocessing into the composite error model to calculate the error compensation value ΔY of the current measurement; the DSP core processing unit generates a compensation signal based on the calculation result, converts the digital signal into an analog signal through a 16-bit DAC conversion chip, amplifies it through a differential amplifier circuit, and connects it to the analog front end through optocoupler isolation to calculate the original measurement result Y of the measuring instrument. raw Perform real-time correction to obtain the compensated measurement result Y compensated =Y raw -ΔY.

[0072] For example, when measuring the voltage of a circuit, the original measurement result Y raw The error compensation value ΔY calculated by the composite error model is 220V, and the compensation result Y is 0.5V. compensated =220-0.5=219.5V.

[0073] 6. Intelligent Optimization of Calibration Parameters (S5)

[0074] Using the least squares method, the instrument regularly analyzes the measurement data and compensation effects; the compensation error threshold is set at 0.5%. When the average compensation error of five consecutive measurements exceeds this threshold, or when trigger conditions such as the ambient temperature change exceeding 5°C or the humidity change exceeding 10% RH are detected, the instrument automatically starts optimizing the coefficients of the composite error model. Based on the current measurement data and actual measurement error, the regression coefficients of the multivariate linear regression model and the connection weights and thresholds of the neural network are recalculated and updated, so that the composite error model can better adapt to changes in the environment and measurement conditions.

[0075] 7. Abnormal Alarm and Handling (S6)

[0076] During the implementation of dynamic error compensation, if the calculated error compensation value ΔY exceeds a reasonable range (e.g., the absolute value is greater than 5% of the original measurement result), the instrument will sound an alarm through a buzzer and display an abnormal prompt message on the display screen, prompting the user to check the instrument and measurement environment so as to promptly discover and resolve potential problems.

[0077] Through real-time scene simulation, the dynamic calibration error compensation method of the electrical measuring instrument of the present invention can effectively deal with the measurement error problem in complex environments, significantly improve the measurement accuracy and reliability of the measuring instrument, and has good practical application value.

[0078] To sum up, the dynamic calibration error compensation method of electrical measuring instruments, the present invention, through real-time acquisition of multi-source data, advanced data preprocessing technology and composite error model construction, combined with multivariate linear regression and improved BP neural network algorithm, can accurately capture the dynamic changes of the environment and measurement signals, effectively filter out interference, and accurately calculate the error compensation value; in actual measurement, such as the 220V voltage measurement, the 0.5V error can be corrected, which greatly improves the measurement accuracy compared with traditional methods, realizes high-precision dynamic measurement, and provides reliable measurement data for industrial production, scientific research and testing and other scenarios.

[0079] In addition, it integrates a variety of high-precision sensors to monitor environmental parameters in real time. When the environmental changes exceed the threshold, the acquisition frequency is automatically increased, and the calibration parameters are dynamically adjusted through an intelligent optimization mechanism, so that the error compensation model can adapt to complex and changing environments. At the same time, the abnormal alarm and processing mechanism, and the precise design of the hardware circuit ensure the stable operation of the system under various working conditions, greatly enhancing the adaptability and reliability of measuring instruments in complex environments, and reducing the risk of equipment failure and maintenance costs.

[0080] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply the existence of any such actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or device comprising the element.

[0081] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A method for compensating for dynamic calibration errors of electrical measuring instruments, characterized in that: The following steps are involved: S1: Real-time acquisition of multi-source data: Integrate temperature, humidity sensors and electromagnetic interference detection modules in the measuring instrument, configure voltage, current, frequency and other measurement channels, and synchronously collect environmental parameters (temperature T, humidity H, electromagnetic interference intensity E) and measurement data (voltage, current, frequency and other model characteristic parameters X1, X2, ..., X) at a frequency of not less than 100HZ. n ); S2: Data preprocessing: Use IIR / FIR digital filter combination to process the measurement signal, improve the signal-to-noise ratio to more than 100dB for 50Hz power frequency interference and harmonic notch; smooth the temperature and humidity data and store the processed data; S3: Construct a composite error model: Use the multivariate linear regression algorithm to build a linear relationship based on the experimental data to form the basic model ΔY linear =f linear (T,H,E,X1,X2,…,X n ); Use machine learning algorithms to mine nonlinear relationships and build a prediction model ΔY nonlinear =f nonlinear (T,H,E,X1,X2,…,X n ), the two are added together to get the composite model ΔY=ΔY linear +ΔY nonlinear ; S4: Dynamic error compensation calculation and implementation: Substitute the processed parameters into the composite model, calculate the compensation value ΔY, and correct the original result Y through the hardware feedback circuit. raw , get Y compensated =Y raw -ΔY; S5: Intelligent optimization of calibration parameters: Uses machine learning algorithms to regularly analyze data and automatically optimize model coefficients when compensation errors exceed 0.5% or environmental changes trigger conditions; S6: Abnormal alarm and processing: If ΔY exceeds the range during compensation, the instrument will sound an alarm and prompt you to check the instrument and environment.

2. The method for compensating for dynamic calibration errors of electrical measuring instruments according to claim 1, characterized in that: The temperature sensor has an accuracy of ±0.1°C, the humidity sensor has an accuracy of ±2%RH, and the electromagnetic interference detection module has an accuracy of ±1dB.

3. The method for compensating for dynamic calibration errors of electrical measuring instruments according to claim 1, wherein: In the multi-source data real-time acquisition step, when the environmental parameters change by more than ±2°C for temperature, ±5%RH for humidity, or ±5dB for electromagnetic interference intensity, high-speed data acquisition is automatically triggered, and the acquisition frequency is increased to 500Hz.

4. The method for compensating for dynamic calibration errors of electrical measuring instruments according to claim 1, wherein: In the IIR / FIR digital filter combination, the order of the IIR filter is 8, the number of taps of the FIR filter is 32, and the notch frequency range is 49 Hz-51 Hz.

5. The method for compensating for dynamic calibration errors of electrical measuring instruments according to claim 1, characterized in that: In the step of constructing the composite error model, after each 1000 dynamic error compensation calculations are performed, the experimental data is updated, and the basic error compensation model and the error prediction model are reconstructed based on the updated data.

6. The method for compensating for dynamic calibration errors of electrical measuring instruments according to claim 1, characterized in that: The machine learning algorithm adopts an improved BP neural network, which includes three hidden layers with 16-32-16 neurons respectively. The activation function is a ReLU-Sigmoid hybrid function. The Adam optimization algorithm is used during training, and the learning rate is dynamically adjusted in the range of 0.001-0.

01.

7. The method for compensating for dynamic calibration errors of electrical measuring instruments according to claim 1, characterized in that: The hardware feedback circuit includes a DSP core processing unit, a 16-bit DAC conversion chip and a differential amplifier circuit. The compensation signal is connected to the analog front end after being isolated by photoelectric coupling, and the compensation delay time is ≤20ms.

8. The method for compensating for dynamic calibration errors of electrical measuring instruments according to claim 1, characterized in that: In the multi-source data real-time acquisition step, the measurement data and environmental parameters are transmitted via the SPI protocol with a transmission rate of 5 Mbps. The data frame contains a 16-bit check code for data integrity verification.

9. The method for compensating for dynamic calibration errors of electrical measuring instruments according to claim 7, characterized in that: The DSP core processing unit, 16-bit DAC conversion chip and differential amplifier circuit in the hardware feedback circuit are synchronized with a unified 100MHz clock source, and the clock jitter is ≤100ps, ensuring the phase consistency of the compensation signal and the measurement signal.

Images