A method and system for real-time control and testing of power frequency magnetic field of electric energy meter

Abstract

The present invention discloses a method and system for testing the real-time control of the industrial frequency magnetic field of an electric energy meter. The method of the present invention comprises the steps of: real-time monitoring of the industrial frequency magnetic field intensity and distribution in the laboratory, and preprocessing of the magnetic field data, capturing the frequency characteristics of the magnetic field signal through the frequency domain multi-scale decomposition algorithm, and using a multi-layer nested adaptive feedback control algorithm on the basis of frequency domain analysis to adjust the output of the magnetic field generator in real time to achieve precise magnetic field control; uploading the test data to the cloud platform, and dynamically adjusting the test scheme using a multi-dimensional fuzzy logic algorithm, generating hidden variables and performing time series prediction through an enhanced time series prediction algorithm, evaluating the performance parameters of the electric energy meter in the industrial frequency magnetic field environment, and providing support for the real-time control test of the electric energy meter. The present invention evaluates the performance parameters of the electric energy meter in the industrial frequency magnetic field environment by analyzing the prediction results, and provides support for the real-time control test of the electric energy meter, so as to achieve the best control of the magnetic field.

Description

Technical Field

[0001] The present invention relates to the technical field of electric energy metering, and in particular to a method and system for testing the real-time control of the industrial frequency magnetic field of an electric energy meter. Background Art

[0002] In actual operation, electric energy meters are often affected by various external environmental factors, among which the interference of power frequency magnetic field is particularly significant. The power frequency magnetic field has a wide range of sources, including power equipment operation, transmission lines, power transformers, etc. These magnetic field interferences may cause measurement errors of electric energy meters and even affect their normal operation. Therefore, how to accurately simulate and regulate the power frequency magnetic field in a laboratory environment and conduct comprehensive and accurate electric energy meter testing has become an important topic of current research.

[0003] At present, the market demand for electric energy meters is growing, and the types and specifications of electric energy meters are also increasing. Electric energy meters of different types and uses have different requirements for test conditions and standards. Traditional test methods and equipment are difficult to meet such diverse and sophisticated test needs. Traditional test methods usually rely on fixed test environments and conditions, lack flexibility and adaptability, and cannot adjust test plans in real time to adapt to the technical specifications and test requirements of different electric energy meters. In addition, traditional test methods have certain limitations in data acquisition, processing and analysis, and it is difficult to make full use of modern digital technology for efficient data processing and in-depth analysis, which limits the accuracy and comprehensiveness of test results.

[0004] The existing technology has at least the following technical problems: lack of multi-scale decomposition means, difficulty in accurately capturing the frequency characteristics of the signal, and inability to identify the frequency components and changing patterns of the magnetic field, resulting in a weak control foundation; inability to dynamically adjust the control gain and prediction error, the system used for real-time control testing of the power frequency magnetic field of the electric energy meter lacks adaptability and accuracy, and the control effect is not ideal; the output of the magnetic field generator is unstable, and the power frequency magnetic field strength and distribution in the laboratory cannot be kept stable and controllable, resulting in insufficient support for fault warning and performance optimization of the electric energy meter. Summary of the invention

[0005] In order to solve the defects of the above-mentioned prior art, the present invention provides a method and system for real-time control and testing of the industrial frequency magnetic field of an electric energy meter, which evaluates the performance parameters of the electric energy meter in the industrial frequency magnetic field environment by analyzing the prediction results, and provides support for the real-time control test of the electric energy meter to achieve optimal control of the magnetic field.

[0006] To this end, the present invention adopts the following technical solution.

[0007] In a first aspect, the present invention provides a method for testing the real-time control of the power frequency magnetic field of an electric energy meter, which comprises the steps of:

[0008] S1. Real-time monitoring of the power frequency magnetic field intensity and distribution data in the laboratory, and pre-processing of the magnetic field data. The frequency characteristics of the magnetic field signal are captured through the frequency domain multi-scale decomposition algorithm. Based on the frequency domain analysis, a multi-layer nested adaptive feedback control algorithm is used to adjust the output of the magnetic field generator in real time to achieve precise magnetic field control.

[0009] S2. Upload the test data to the cloud platform, and use the multi-dimensional fuzzy logic algorithm to dynamically adjust the test plan. Through the enhanced timing prediction algorithm, generate hidden variables and perform timing prediction to evaluate the performance parameters of the electric energy meter in the industrial frequency magnetic field environment, and provide support for the real-time control test of the electric energy meter.

[0010] Furthermore, S1 specifically includes: converting the collected raw data into a digital signal through an analog-to-digital converter inside the magnetic field sensor, performing data preprocessing on the converted digital signal, including denoising, filtering and data calibration, processing the calibrated magnetic field data through a frequency domain multi-scale decomposition algorithm, and decomposing the time domain signal into multiple frequency domain signal components.

[0011] Furthermore, in S1, the formula used by the frequency domain multi-scale decomposition algorithm is as follows:

[0012]

[0013] in, is the nth frequency component, which represents the signal strength of the magnetic field signal in the frequency domain; Indicates frequency; is the signal strength at the kth time sampling point; is the total number of sampling points; ; Indicates time; is the index of the frequency component.

[0014] Furthermore, in S1, the multi-layer nested adaptive feedback control algorithm is divided into three layers: the first layer is a frequency domain adaptive controller, which dynamically adjusts the control gain according to the frequency domain analysis results of the frequency domain multi-scale decomposition algorithm to obtain a frequency domain control signal; the second layer is a time domain prediction controller, which performs predictive adjustment through a time domain prediction model on the basis of the frequency domain control signal of the first layer to obtain a time domain control signal; the third layer is a combined optimization controller, which combines the frequency domain and time domain control signals, optimizes the control signal through a genetic algorithm, and ensures global optimization.

[0015] Furthermore, in S2, mining and pattern recognition are performed through an enhanced time series prediction algorithm, and the enhanced time series prediction algorithm combines a deep convolutional variational autoencoder with a multilayer perceptron to perform convolution and encoding operations on the input test data, and the convolved and encoded data generates latent variables through a variational autoencoder.

[0016] Furthermore, in S2, the latent variables are subjected to adaptive multi-scale transformation to capture the multi-scale characteristics of the data and the multi-scale dependencies between the latent variables, and nonlinear mapping is performed on the latent variables after the multi-scale transformation to further extract their features.

[0017] Furthermore, in S2, in combination with the nonlinear mapping output, data of different scales are aligned through dynamic time warping to solve the time alignment problem between data of different scales. By combining the state transfer and observation model, the data after dynamic time warping is predicted in time series to generate predictions for future moments.

[0018] Furthermore, in S2, nonlinear mapping and dynamic time warping outputs are integrated to generate timing prediction results, thereby ensuring accurate capture of complex timing dependencies and nonlinear characteristics; and corresponding magnetic field control strategies are generated based on the timing prediction results.

[0019] Furthermore, in S2, the magnetic field control strategy includes:

[0020] Adjust the magnetic field strength: increase or decrease the magnetic field strength according to the predicted performance parameters of the electric energy meter to ensure the stability and accuracy of the electric energy meter;

[0021] Adjust the frequency of the magnetic field: According to the response time and fault prediction of the electric energy meter, adjust the frequency of the magnetic field to avoid abnormalities of the electric energy meter at a specific frequency;

[0022] Optimize magnetic field distribution: According to the response of the electric energy meter at different positions, adjust the distribution of the power frequency magnetic field in the laboratory to ensure that each electric energy meter is evenly stressed.

[0023] In a second aspect, the present invention provides a power meter power frequency magnetic field real-time control test system, which is used for the above power meter power frequency magnetic field real-time control test method, and comprises:

[0024] Magnetic field control unit: monitors the power frequency magnetic field intensity and distribution data in the laboratory in real time, and performs magnetic field data preprocessing. It captures the frequency characteristics of the magnetic field signal through the frequency domain multi-scale decomposition algorithm. Based on the frequency domain analysis, it uses a multi-layer nested adaptive feedback control algorithm to adjust the output of the magnetic field generator in real time to achieve precise magnetic field control.

[0025] Timing prediction unit: uploads the test data to the cloud platform, and uses a multi-dimensional fuzzy logic algorithm to dynamically adjust the test plan. Through the enhanced timing prediction algorithm, it generates hidden variables and performs timing prediction to evaluate the performance parameters of the electric energy meter in the industrial frequency magnetic field environment, providing support for the real-time control test of the electric energy meter.

[0026] The present invention has the following beneficial effects:

[0027] 1. Through the frequency domain multi-scale decomposition algorithm, the time domain signal is decomposed into multiple frequency domain signal components, which can more accurately capture the frequency characteristics of the signal, identify the frequency components and change laws of the magnetic field, and provide a reliable foundation for subsequent magnetic field regulation; a multi-layer nested adaptive feedback control algorithm is designed, which is divided into three layers: frequency domain adaptive controller, time domain predictive controller and combined optimization controller. The frequency domain adaptive controller dynamically adjusts the control gain, and the time domain predictive controller improves the predictability and stability of the control signal by introducing the prediction error for a period of time in the future. The combined optimization controller ensures the global optimum through the genetic algorithm, thereby achieving the best regulation of the magnetic field.

[0028] 2. Through the multi-layer nested adaptive feedback control algorithm, the magnetic field generator can adjust the output in real time to ensure that the power frequency magnetic field strength and distribution in the laboratory remain stable and controllable. All test data are uploaded to the cloud platform in real time via the wireless network. The multi-dimensional fuzzy logic algorithm is used to dynamically adjust the test plan according to the technical specifications and test requirements of the electric energy meter to ensure the comprehensiveness and accuracy of the test; on the cloud platform, the test data is deeply mined and pattern recognized through the enhanced time series prediction algorithm. Combining the deep convolutional variational autoencoder with multi-scale nonlinear mapping and dynamic time warping algorithm, it can accurately capture complex time series dependencies and nonlinear characteristics and generate predictions for future moments. By analyzing the prediction results, the abnormal conditions or performance degradation trends that may occur in the electric energy meter can be identified.

[0029] 3. The present invention not only improves the accuracy and comprehensiveness of the electric energy meter test, but also enhances the support for fault warning and performance optimization of the electric energy meter. From magnetic field monitoring and regulation to data uploading and processing, a complete closed loop is formed, which effectively solves the limitations of traditional testing methods and provides important guarantees for the reliable operation of the electric energy meter; through multi-dimensional fuzzy logic and enhanced timing prediction algorithms, the performance of the electric energy meter in the power frequency magnetic field environment can be comprehensively evaluated, including indicators such as the reading accuracy, stability, failure rate, response time and energy consumption of the electric energy meter. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 This is a flow chart of a method for real-time control of the power frequency magnetic field of an electric energy meter according to the present invention;

[0031] Figure 2 The present invention is a structural diagram of a real-time control and testing system for an electric energy meter's industrial frequency magnetic field. DETAILED DESCRIPTION

[0032] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the technical scheme in the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiment of the present invention. Obviously, the described embodiment is only a part of the embodiment of the present invention, not all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0034] Example 1

[0035] Refer to the attached Figure 1 , which shows a flow chart of a method for real-time control of power frequency magnetic field of an electric energy meter provided by an embodiment of the present invention, the steps of the method are as follows:

[0036] S1. Real-time monitoring of the industrial frequency magnetic field intensity and distribution data in the laboratory, and pre-processing of the magnetic field data. The frequency characteristics of the magnetic field signal are captured through the frequency domain multi-scale decomposition algorithm. Based on the frequency domain analysis, a multi-layer nested adaptive feedback control algorithm is used to adjust the output of the magnetic field generator in real time to achieve precise magnetic field control.

[0037] Magnetic field sensors (such as Hall effect sensors) are used to monitor the intensity and distribution of the power frequency magnetic field in the laboratory in real time. Magnetic field sensors are placed in key locations in the laboratory to ensure comprehensive coverage of the entire environmental magnetic field. In order to achieve high-precision and high-frequency data acquisition, the sampling frequency of the magnetic field sensor is set to several thousand hertz, which can capture rapidly changing magnetic field signals and ensure the timeliness and accuracy of the data. The magnetic field signal is the data of the change of magnetic field intensity over time, which is a set of time series data. Therefore, the collected magnetic field signal itself is a time domain signal.

[0038] The collected raw data (the magnetic field data composed of the industrial frequency magnetic field strength and distribution data, i.e., raw data) is converted into digital signals by the analog-to-digital converter (ADC) inside the sensor, and the converted data is preprocessed, including denoising, filtering, and data calibration. Denoising processing uses digital filters (such as Kalman filters) to reduce noise in the data and improve the signal-to-noise ratio of the signal. Filtering processing uses bandpass filters to extract the frequency components of interest and exclude irrelevant frequency interference. Data calibration corrects the collected data based on known reference signals to ensure the accuracy of the measurement results.

[0039] The calibrated magnetic field data is processed by a frequency domain multi-scale decomposition algorithm, which decomposes the time domain signal into multiple frequency domain signal components to more accurately capture the frequency characteristics of the signal. The specific formula is as follows:

[0040]

[0041] in, is the nth frequency component, indicating the corresponding signal strength in the frequency domain, Indicates frequency, is the signal strength at the kth time sampling point, is the total number of sampling points, , Indicates time, It is the index of the frequency component. Through multi-scale analysis in the frequency domain, the frequency components and changing patterns of the magnetic field can be identified, providing a basis for subsequent regulation.

[0042] Based on the frequency domain analysis, a multi-layer nested adaptive feedback control algorithm is used for regulation. The target magnetic field strength is set to , the real-time magnetic field strength is , then the control error for:

[0043]

[0044] The multi-layer nested adaptive feedback control algorithm is divided into three layers, each of which has a unique innovative design to ensure high adaptability and accuracy in the implementation of the control test process. The first layer is a frequency domain adaptive controller, which dynamically adjusts the control gain according to the frequency domain analysis results of the frequency domain multi-scale decomposition algorithm. Its formula is as follows:

[0045]

[0046] in, is the first-layer frequency domain control signal, is the frequency domain proportional control gain parameter, is the frequency domain integral control gain parameter, is the frequency domain differential control gain parameter, is the frequency domain error, It is a time-integrated variable. By adjusting the frequency domain gain parameters, it can dynamically respond to changes in different frequency components and enhance control accuracy.

[0047] The second layer is the time domain prediction controller, which performs prediction and adjustment through the time domain prediction model based on the frequency domain control signal of the first layer. The formula is as follows:

[0048]

[0049] in, is the second layer time domain control signal, is the time step, is the time domain proportional control gain parameter, is the time domain integral control gain parameter, is the time domain derivative control gain parameter, In the future time The prediction error, It is the future time The second-layer time-domain predictive controller improves the predictability and stability of the control signal by introducing the prediction error for a period of time in the future, ensuring a more accurate response in the future.

[0050] The third layer is a combined optimization controller, which combines the frequency domain and time domain control signals and optimizes the control signals through genetic algorithms to ensure global optimization. The specific formula is as follows:

[0051]

[0052] in, is the final combined optimized control signal, It is a genetic algorithm used to optimize the input signal. The genetic algorithm is a prior art and the present invention will not elaborate on it in detail. The genetic algorithm continuously optimizes the control signal through selection, crossover and mutation operations. The selection operation selects individuals with high fitness through the evaluation function; the crossover operation generates new individuals by exchanging gene information; the mutation operation increases the population diversity by randomly changing the gene information. This ensures that the control signal finally output is the global optimal solution, and achieves the best regulation of the magnetic field.

[0053] The magnetic field generator is based on the final control signal Adjust the output in real time to ensure that the power frequency magnetic field strength and distribution in the laboratory remain stable and controllable. The new magnetic field strength is recorded as:

[0054]

[0055] in, The frequency domain decomposition algorithm and the multi-layer feedback control algorithm are closely combined to achieve precise control of the industrial frequency magnetic field.

[0056] S2. Upload the test data to the cloud platform, and use the multi-dimensional fuzzy logic algorithm to dynamically adjust the test plan. Through the enhanced timing prediction algorithm, generate hidden variables and perform timing prediction to evaluate the performance parameters of the electric energy meter in the industrial frequency magnetic field environment, and provide support for the real-time control test of the electric energy meter.

[0057] The test data includes a timestamp, a new magnetic field strength value, and performance parameters of the electric energy meter, wherein the new magnetic field strength value is the final output of S1.

[0058] All test data are uploaded to the cloud platform in real time via wireless network. The data format is ,in is the i-th timestamp, is the i-th new magnetic field strength value, recorded in The magnetic field strength at the moment, is the performance parameter of the ith electric energy meter, recorded in The performance data of the electric energy meter at the moment. Using the multidimensional fuzzy logic algorithm, the test plan is dynamically adjusted according to the technical specifications and test requirements of the electric energy meter. The multidimensional fuzzy logic rule base is defined as , each rule is of the form:

[0059]

[0060] in, is the input variable, is the membership function, is the output variable, The output membership function is obtained by using fuzzy logic algorithm to dynamically adjust the test plan under different test conditions to ensure the comprehensiveness and accuracy of the test.

[0061] On the cloud platform, the test data is mined and pattern recognized by an enhanced time series prediction algorithm, which combines a deep convolutional variational autoencoder with a multi-layer perceptron to perform convolution and encoding operations on the input test data:

[0062]

[0063] in, is the output after convolution and encoding operations, is the encoder weight, is the input test data, , is the activation function, is the bias, The convolution operation can extract the local features of the input test data, and the encoding operation compresses the local features into a low-dimensional representation for subsequent processing.

[0064] The convolved and encoded data is passed through a variational autoencoder to generate latent variables:

[0065]

[0066] in, is a hidden variable, is the conditional distribution, is a Gaussian distribution, is the mean of the latent variable, is the variance of the latent variable, Represents Gaussian distribution. The variational autoencoder reconstructs the input data so that the generated latent variables can better represent the potential structure of the data.

[0067] The latent variables are transformed through adaptive multi-scale transformation to capture the multi-scale characteristics of the data and the multi-scale dependencies between the latent variables. The formula is as follows:

[0068]

[0069] in, is the time t The latent variable after scale transformation, For the The scale of high-order adaptive transform coefficients, is the adaptive transformation order, is the time t The noise term at this scale, It is Hidden variables at time.

[0070] The latent variables after multi-scale transformation are processed by nonlinear mapping to further extract their features. The mapping model is:

[0071]

[0072] in, is the time t The hidden state after nonlinear mapping at the same scale, For the The weight matrix under scale, For the Bias vector at scale.

[0073] Combined with the nonlinear mapping output, data of different scales are aligned through time warping to solve the time alignment problem between data of different scales. The formula is as follows:

[0074]

[0075]

[0076] in, It is in Scale Time and The distance metric after the moments are aligned by dynamic time warping, is the time t The final distance metric after dynamic time warping alignment at scale, is the time warp weight, indicating the Time and The alignment weight at the moment, and Respectively represent Time and Moment Hidden state after nonlinear mapping at different scales; represents the Euclidean distance.

[0077] By combining the state transfer and observation models, the dynamic time warped data is used for time series prediction to generate predictions for future moments. The formula is as follows:

[0078]

[0079] in, is the state of the test system at time t, is the state transfer matrix, is the control input matrix, is the control input, i.e. the final combined optimized control signal; is the process noise at time t, Represents the data after dynamic time warping alignment at time t−1.

[0080] Fusion of nonlinear mapping and dynamic time warping outputs to generate time series prediction results ensures accurate capture of complex time series dependencies and nonlinear features:

[0081]

[0082] in, It is the time series prediction result at time t, predicting the performance parameters or status of the electric energy meter in the power frequency magnetic field environment, including the reading accuracy, stability, failure rate, response time, energy consumption, etc. of the electric energy meter. is the observation matrix, which is used to map the state space to the observation space. For the The fusion weight of the scale, To output the bias vector, adjust the offset of the prediction result.

[0083] According to the time series prediction results, a corresponding magnetic field control strategy is generated to ensure that the electric energy meter operates under the best working conditions. The magnetic field control strategy is as follows:

[0084] Adjust the magnetic field strength: increase or decrease the magnetic field strength according to the predicted performance parameters of the electric energy meter to ensure the stability and accuracy of the electric energy meter;

[0085] Adjust the frequency of the magnetic field: According to the response time and fault prediction of the electric energy meter, adjust the frequency of the magnetic field to avoid abnormalities of the electric energy meter at a specific frequency;

[0086] Optimize magnetic field distribution: According to the response of the electric energy meter at different positions, adjust the distribution of the magnetic field in the laboratory to ensure that each electric energy meter is evenly stressed.

[0087] In summary, a real-time control test method for the industrial frequency magnetic field of an electric energy meter has been completed.

[0088] The above-mentioned power meter power frequency magnetic field real-time control test method is given as an example with data as follows.

[0089] 1. Experimental environment and data collection

[0090] Assume that multiple Hall effect sensors are used in the laboratory and arranged at key locations to monitor the intensity and distribution of the power frequency magnetic field in the laboratory in real time. In order to ensure comprehensive coverage of the entire environmental magnetic field, the sampling frequency of the sensor is set to 5000Hz, and the sensor is arranged in an area with large magnetic field changes.

[0091] The collected data is the relationship between the power frequency magnetic field strength and time. Assume that the following magnetic field data (unit: ):

[0092]

[0093] Data collection interval: .

[0094] The data includes the magnetic field strength at 10 moments, assuming the sampling time is 50ms.

[0095] Target magnetic field strength: The target is set to .

[0096] 2. Data Preprocessing

[0097] State estimation after Kalman filter denoising: .

[0098] Since the power frequency magnetic field signal is periodically changing (50Hz), a bandpass filter is used to extract the frequency components around 50Hz. Assume that the center frequency of the bandpass filter is 50Hz and the bandwidth is 10Hz.

[0099] The filtered signal mainly contains the frequency component of 50±5Hz, and other frequencies (such as 100Hz or 30Hz, etc.) are filtered out, thereby ensuring that an accurate power frequency signal is obtained.

[0100] Assuming the standard The reference signal is used to calibrate the data. Using a standard reference signal, all sensor readings are corrected to accurately represent the actual magnetic field strength. The calibration process assumes a linear correction for all acquired data.

[0101] 3. Frequency domain analysis and decomposition

[0102] The frequency domain of the data is decomposed by Fourier transform to identify the frequency characteristics of the signal. The discrete Fourier transform (DFT) formula is used to analyze the frequency domain components of the signal.

[0103]

[0104] For the magnetic field data sampled in this example , and perform discrete Fourier transform on it.

[0105] 3.1 Fourier transform calculation process

[0106] Assume that magnetic field data has been collected for 10 moments:

[0107] , and the sampling period is , there are 10 sample points in total.

[0108] Perform Fourier transform on these data and calculate their frequency components.

[0109] For example, to calculate the first frequency component :

[0110] because ,so:

[0111]

[0112] The same process applies to other frequency components, such as , Etc., by calculating each frequency component, the complete frequency domain signal is obtained.

[0113] 3.2 Frequency Component Identification

[0114] After calculating the frequency components after Fourier transformation, the frequency components of the signal can be identified. For example, there may be a clear peak around 50Hz, indicating that the signal has power frequency characteristics.

[0115] 4. Multi-layer nested adaptive feedback control algorithm

[0116] By using a multi-layer nested adaptive feedback control algorithm to accurately control the strength of the magnetic field, the electric energy meter can be ensured to work stably in a complex magnetic field environment. In this embodiment, the control algorithm is divided into three layers: the first layer is frequency domain feedback control, the second layer is time domain control, and the third layer is combined optimization control. The calculation of each layer of control signal depends on the output of the previous layer and gradually optimizes the result.

[0117] 4.1 First-layer frequency domain control signal calculation

[0118] In the first level of frequency domain control, the control signal is calculated based on the error between the real-time measured magnetic field signal and the target magnetic field value. Assume that the target magnetic field strength is , we use the following control rules:

[0119]

[0120] in, is the frequency domain error, is the frequency component currently measured.

[0121] The magnetic field signal obtained from data acquisition is:

[0122]

[0123] Use Discrete Fourier Transform (DFT) to convert the sampled data from the time domain to the frequency domain. Assume that the data has been analyzed by DFT and the frequency components of the signal have been obtained. For example, assume that the frequency component around 50Hz is:

[0124]

[0125] Calculate the frequency domain error:

[0126]

[0127] Use proportional control to calculate the first-level control signal, assuming the proportional gain ,but:

[0128]

[0129] Control Signal The output of the first layer as feedback is passed to the second layer of control.

[0130] 4.2 Second-layer time domain control signal calculation

[0131] In the second level of time-domain control, the time-domain error and the target control signal are used to further adjust the magnetic field. The goal of time-domain control is to dynamically adjust the magnetic field to eliminate the persistent error.

[0132] The error is calculated through continuous magnetic field measurement data. Assume that at a certain moment, the currently measured magnetic field is , the error is:

[0133]

[0134] For example, at a certain moment The magnetic field strength is , then the time domain error is:

[0135]

[0136] Integrate the time domain error to obtain the cumulative value of the error, assuming that the time interval The cumulative error is ;

[0137] Assume that the error sequence during this period is , respectively:

[0138]

[0139] The accumulated error is:

[0140] Calculate the time domain control signal, assuming that the proportional gain of the time domain control is , the integral gain is , (Current moment error), then the second layer control signal is:

[0141]

[0142] The second layer of control signal The output of the second layer feedback is passed to the third layer for optimization control.

[0143] 4.3 Third-layer combined optimization control signal calculation

[0144] In the third layer, combined optimization control is adopted to combine the outputs of frequency domain control and time domain control, and the control signal is further optimized through genetic algorithm (GA) or other global optimization methods to ensure system stability and fast response.

[0145] Output the first layer Second layer output Perform weighted synthesis to form the final control signal. Assume that the weighting factor is: ,in, The weight of is 0.8, The weight of is 0.2.

[0146] Calculate the weighted combined control signal:

[0147]

[0148] At this time, the final control signal is obtained To further optimize the signal, global optimization methods such as genetic algorithm (GA) or particle swarm optimization (PSO) can be used to optimize the Fine adjustments were made to ensure stability of the magnetic field strength throughout the experiment.

[0149] 4.4 Feedback Regulation and Experimental Results

[0150] Through the output of the above three-layer nested feedback control algorithm, the power frequency magnetic field strength in the test environment of the electric energy meter can be adjusted. During the experiment, the accuracy of the electric energy meter is monitored in real time, and the control signal is dynamically adjusted according to the change of the magnetic field. Adaptive adjustment is continued based on the error data fed back.

[0151] Experimental data:

[0152] After a series of experiments, the control signal Effectively maintains magnetic field strength close to target value For example, in the 10th experiment, the measured magnetic field strength is , the error is very small, indicating that the control system has achieved a high accuracy.

[0153] Experimental conclusion:

[0154] 1. Through the multi-layer nested adaptive feedback control algorithm, the magnetic field strength can be accurately adjusted to the target value .

[0155] 2. The control system shows good stability and adaptability under multiple feedback adjustments, ensuring the accuracy of the performance test of the electric energy meter in a complex magnetic field environment.

[0156] 3. Further experiments show that the algorithm can maintain high control accuracy and reliability under different magnetic field strengths and interference conditions.

[0157] In the implementation plan, by real-time monitoring of magnetic field signals and applying frequency domain multi-scale decomposition algorithm and multi-layer nested adaptive feedback control algorithm, the final control signal can accurately adjust the strength of the magnetic field and ensure that the magnetic field strength in the laboratory is close to the target value. For example, during multiple experiments, the magnetic field strength adjusted by the algorithm was able to remain close to the target value:

[0158] In the 10th experiment, the measured magnetic field strength is , the error is less than .

[0159] In other experiments, the fluctuation range of magnetic field strength was controlled within It meets the experimental accuracy requirements.

[0160] This result shows that compared with traditional magnetic field control methods (such as simple PID control or control strategies based on static calibration), this scheme can achieve more accurate and real-time control in dynamic and changing environments. The stability of the magnetic field is greatly improved, providing a more reliable experimental environment for accurate testing of electric energy meters.

[0161] Example 2

[0162] This embodiment provides a power meter power frequency magnetic field real-time control test system, which is used for the power meter power frequency magnetic field real-time control test method described in the above embodiment 1, such as Figure 2 As shown, it includes:

[0163] Magnetic field control unit: monitors the power frequency magnetic field intensity and distribution data in the laboratory in real time, and performs magnetic field data preprocessing. It captures the frequency characteristics of the magnetic field signal through the frequency domain multi-scale decomposition algorithm. Based on the frequency domain analysis, it uses a multi-layer nested adaptive feedback control algorithm to adjust the output of the magnetic field generator in real time to achieve precise magnetic field control.

[0164] Timing prediction unit: uploads the test data to the cloud platform, and uses a multi-dimensional fuzzy logic algorithm to dynamically adjust the test plan. Through the enhanced timing prediction algorithm, it generates hidden variables and performs timing prediction to evaluate the performance parameters of the electric energy meter in the industrial frequency magnetic field environment, providing support for the real-time control test of the electric energy meter.

[0165] It should be noted that each unit in the above-mentioned power meter power frequency magnetic field real-time control test system can be fully or partially implemented by software, hardware and their combination. The above-mentioned units can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above-mentioned units. For the specific definition of a power meter power frequency magnetic field real-time control test system, please refer to the definition of a power meter power frequency magnetic field real-time control test method mentioned above. The two have the same functions and effects, which will not be repeated here.

[0166] The sequence of the embodiments of the present invention is for description only and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0167] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0168] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention.

Claims

1. A method for real-time control and testing of power frequency magnetic field of an electric energy meter, characterized in that: Includes steps: S1. Real-time monitoring of the power frequency magnetic field intensity and distribution data in the laboratory, and pre-processing of the magnetic field data. The frequency characteristics of the magnetic field signal are captured through the frequency domain multi-scale decomposition algorithm. Based on the frequency domain analysis, a multi-layer nested adaptive feedback control algorithm is used to adjust the output of the magnetic field generator in real time to achieve precise magnetic field control. S2. Upload the test data to the cloud platform and dynamically adjust the test plan using a multi-dimensional fuzzy logic algorithm. Generate latent variables and perform time series prediction using an enhanced time series prediction algorithm to evaluate the performance parameters of the electric energy meter in an industrial frequency magnetic field environment, providing support for real-time control testing of the electric energy meter. In S1, the formula used by the frequency domain multi-scale decomposition algorithm is as follows: ， in, is the nth frequency component, which represents the signal strength of the magnetic field signal in the frequency domain; Indicates frequency; is the signal strength at the kth time sampling point; is the total number of sampling points; ; Indicates time; is the index of the frequency component; In S2, mining and pattern recognition are performed by an enhanced time series prediction algorithm, which combines a deep convolutional variational autoencoder with a multi-layer perceptron to perform convolution and encoding operations on the input test data, and the convolved and encoded data is used to generate hidden variables through a variational autoencoder; the hidden variables are subjected to an adaptive multi-scale transformation to capture the multi-scale characteristics of the data and the multi-scale dependencies between the hidden variables, and the hidden variables after the multi-scale transformation are subjected to nonlinear mapping processing to further extract their features; In S2, the data of different scales are aligned by dynamic time warping in combination with the nonlinear mapping output, so as to solve the time alignment problem between the data of different scales. The data after dynamic time warping is predicted in time series by combining the state transfer and observation model to generate predictions for future moments. In S2, nonlinear mapping and dynamic time warping output are integrated to generate a time series prediction result, ensuring accurate capture of complex time series dependencies and nonlinear characteristics; according to the time series prediction result, a corresponding magnetic field control strategy is generated; the magnetic field control strategy includes: Adjust the magnetic field strength: increase or decrease the magnetic field strength according to the predicted performance parameters of the electric energy meter to ensure the stability and accuracy of the electric energy meter; Adjust the frequency of the magnetic field: According to the response time and fault prediction of the electric energy meter, adjust the frequency of the magnetic field to avoid abnormalities of the electric energy meter at a specific frequency; Optimize magnetic field distribution: According to the response of the electric energy meter at different positions, adjust the distribution of the power frequency magnetic field in the laboratory to ensure that each electric energy meter is evenly stressed.

2. The method for real-time control of the power frequency magnetic field of an electric energy meter according to claim 1 is characterized in that: The S1 specifically includes: converting the collected raw data into a digital signal through an analog-to-digital converter inside the magnetic field sensor, performing data preprocessing on the converted digital signal, including denoising, filtering and data calibration, processing the calibrated magnetic field data through a frequency domain multi-scale decomposition algorithm, and decomposing the time domain signal into multiple frequency domain signal components.

3. The method for real-time control and testing of power frequency magnetic field of electric energy meter according to claim 1 is characterized in that: In S1, The multi-layer nested adaptive feedback control algorithm is divided into three layers: the first layer is a frequency domain adaptive controller, which dynamically adjusts the control gain according to the frequency domain analysis results of the frequency domain multi-scale decomposition algorithm to obtain a frequency domain control signal; the second layer is a time domain prediction controller, which performs prediction and adjustment through a time domain prediction model on the basis of the frequency domain control signal of the first layer to obtain a time domain control signal; the third layer is a combined optimization controller, which combines the frequency domain and time domain control signals, optimizes the control signal through a genetic algorithm, and ensures global optimization.

4. A power meter power frequency magnetic field real-time control test system, which is used to execute the power meter power frequency magnetic field real-time control test method according to any one of claims 1 to 3, characterized in that: include: Magnetic field control unit: monitors the power frequency magnetic field intensity and distribution data in the laboratory in real time, and performs magnetic field data preprocessing. It captures the frequency characteristics of the magnetic field signal through the frequency domain multi-scale decomposition algorithm. Based on the frequency domain analysis, it uses a multi-layer nested adaptive feedback control algorithm to adjust the output of the magnetic field generator in real time to achieve precise magnetic field control. Timing prediction unit: uploads the test data to the cloud platform, and uses a multi-dimensional fuzzy logic algorithm to dynamically adjust the test plan. Through the enhanced timing prediction algorithm, it generates hidden variables and performs timing prediction to evaluate the performance parameters of the electric energy meter in the industrial frequency magnetic field environment, providing support for the real-time control test of the electric energy meter.

Images