A virtual sensing method and system for the moisture content inside a solid insulation material

Abstract

The present invention relates to the technical field of data detection, and relates to a virtual sensing method and system for the internal moisture content of a solid insulation material, including the following steps: obtaining characteristic parameters of the solid insulation material; constructing a virtual detection environment model according to the characteristic parameters of the solid insulation material; applying a simulated electric field in the virtual detection environment model, and obtaining the change data of the polarization current and the depolarization current; performing digital processing on the change data of the polarization current and the depolarization current to generate corresponding spectrum data; performing numerical calculation according to the spectrum data and a preset objective function to obtain a prediction result of the internal moisture content of the solid insulation material; comparing the prediction result with the actually detected moisture content data to obtain an evaluation result; the present invention can perform prediction through a virtual detection environment model, without physically damaging the solid insulation material, ensuring the integrity and reusability of the material.

Description

Technical Field

[0001] The present invention relates to the technical field of data detection, and more specifically, to a virtual sensing method and system for the moisture content inside a solid insulating material. Background Art

[0002] In power systems and electrical equipment, solid insulating materials play a crucial role, and their performance directly affects the safe operation and service life of the equipment. However, solid insulating materials are prone to being affected by various factors during operation, and the moisture content is a key factor. Moisture not only reduces the electrical performance of the insulating material but may also cause equipment failures and even lead to safety accidents.

[0003] Traditional moisture content detection methods sometimes still require physical sampling and destructive testing of the insulating material, which is not only time-consuming and laborious but also unable to monitor the internal moisture changes of the material in real time. In addition, these methods may not be able to comprehensively reflect the internal moisture distribution of the insulating material, so there are certain limitations in practical applications. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to overcome the deficiencies of the prior art and provide a virtual sensing method and system for the moisture content inside a solid insulating material, which can be predicted through a virtual detection environment model without physically damaging the solid insulating material, ensuring the integrity and reusability of the material.

[0005] The basic concept of the technical solution adopted by the present invention to solve the above technical problem is as follows:

[0006] In a first aspect, a virtual sensing method for the moisture content inside a solid insulating material includes:

[0007] Obtain the characteristic parameters of the solid insulating material;

[0008] Construct a virtual detection environment model according to the characteristic parameters of the solid insulating material;

[0009] Apply a simulated electric field in the virtual detection environment model and obtain the change data of the polarization current and depolarization current;

[0010] Digitally process the change data of the polarization current and depolarization current to generate corresponding spectral data;

[0011] Perform numerical calculations according to the spectral data and a preset objective function to obtain the prediction result of the moisture content inside the solid insulating material;

[0012] Compare the prediction result with the moisture content data detected actually to obtain an evaluation result.

[0013] Further, obtain the characteristic parameters of the solid insulation material, including:

[0014] Obtain the original data on the characteristics of the solid insulation material;

[0015] Extract the key features related to the characteristics of the solid insulation material from the original data;

[0016] Cluster the key features, calculate the distance of each data point to each cluster center, and assign it to the corresponding cluster center; for each cluster, calculate the mean of all data points inside it, and use this mean as the new cluster center; repeat the steps until the preset number of iterations is reached. After clustering, different combinations of material characteristics will be obtained, and each category represents a group of materials with similar characteristics.

[0017] Further, the original data includes the resistivity, dielectric constant, breakdown voltage, and heat resistance of the material; the key features include the chemical composition, physical structure, and processing technology of the material.

[0018] Further, digitally process the change data of the polarization current and depolarization current to generate the corresponding spectral data, including:

[0019] Periodically obtain the electrical signal values according to the change data of the polarization current and depolarization current, and convert the electrical signal values into a finite number of discrete digital signal values;

[0020] Reconstruct the polarization current and depolarization current data in the time domain according to the discrete digital signal values;

[0021] Calculate the corresponding spectral data according to the polarization current and depolarization current data in the time domain.

[0022] Further, periodically obtain the electrical signal values according to the change data of the polarization current and depolarization current, and convert the electrical signal values into a finite number of discrete digital signal values, including:

[0023] Determine the sampling frequency and set the sampling time length;

[0024] Sample the polarization current and depolarization current according to the set sampling frequency and time length;

[0025] Each sampling obtains an analog electrical signal value representing the current magnitude at the corresponding time point;

[0026] Send the analog electrical signal value into an analog-to-digital converter, and the analog-to-digital converter maps each analog electrical signal value to a corresponding number to obtain the discrete digital signal value.

[0027] Further, the analog-to-digital converter maps each analog electrical signal value to a corresponding number, including:

[0028] Determine the range of the analog electrical signal and the range of the digital signal. Let the minimum value of the analog electrical signal be a min , and the maximum value be a max ;

[0029] Based on the minimum value of the analog electrical signal being a min and the maximum value being a max , calculate the range width a of the analog signal r = a max − a min ;

[0030] Calculate the range width of the digital signal, and based on the range width of the digital signal, calculate the mapping ratio

[0031] For each analog electrical signal value a, perform the following steps to map it to the corresponding digital signal value:

[0032] Normalize the analog electrical signal value a within the range through n o = a − a min , where n o represents the normalized value;

[0033] Map the normalized value to within the digital signal range using the scale factor through d v = n o × s f , and d v represents the corresponding digital value after mapping.

[0034] Furthermore, perform numerical calculations based on the spectral data and a preset objective function to obtain the prediction result of the moisture content inside the solid insulating material, including:

[0035] Analyze the spectral data to obtain the distribution of amplitude with frequency;

[0036] Define the objective function for the relationship between the spectral data and the moisture content based on historical data;

[0037] Based on the spectral data and the objective function, obtain the prediction result of the moisture content inside the solid insulating material.

[0038] In a second aspect, a virtual sensing system for the moisture content inside a solid insulating material includes:

[0039] An acquisition module, configured to acquire the characteristic parameters of the solid insulating material; construct a virtual detection environment model based on the characteristic parameters of the solid insulating material; apply an analog electric field in the virtual detection environment model, and acquire the change data of the polarization current and the depolarization current.

[0040] A processing module for digitally processing the change data of the polarization current and the depolarization current to generate corresponding spectral data; performing numerical calculations based on the spectral data and a preset objective function to obtain a prediction result of the moisture content inside the solid insulating material; comparing the prediction result with the actually detected moisture content data to obtain an evaluation result.

[0041] After adopting the above technical solution, the present invention has the following beneficial effects compared with the prior art: Prediction is carried out through a virtual detection environment model, without physically damaging the solid insulating material, ensuring the integrity and reusability of the material; By applying a simulated electric field and obtaining the change data of the polarization current and the depolarization current, information about the moisture content inside the insulating material can be efficiently obtained; Numerical calculations are carried out using a preset objective function to predict the moisture content inside the solid insulating material, which helps to timely discover potential moisture problems and take corresponding preventive measures; The virtual sensing method reduces the frequency and need for physical detection, thereby reducing the maintenance cost and labor cost, and can predict and handle potential problems in advance; By accurately detecting the moisture content inside the insulating material, potential safety hazards can be timely discovered and handled, thereby improving the overall safety of the electrical system.

[0042] The following further describes in detail the specific embodiments of the present invention with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The accompanying drawings described herein are used to provide a further understanding of the present application, and constitute a part of the present application. The illustrative embodiments and descriptions thereof of the present application are used to explain the present application, and do not constitute an improper limitation of the present application. Some specific embodiments of the present application will be described in detail hereinafter with reference to the accompanying drawings in an exemplary rather than restrictive manner. The same reference numerals in the drawings denote the same or similar components or parts. Those skilled in the art should understand that these drawings are not necessarily drawn to scale. In the drawings:

[0044] Figure 1 is a schematic flowchart of a virtual sensing method for the moisture content inside a solid insulating material of the present invention.

[0045] Figure 2 is a schematic diagram of a virtual sensing system for the moisture content inside a solid insulating material of the present invention.

[0046] It should be noted that these drawings and the text description are not intended to limit the scope of the concept of the present invention in any way, but to illustrate the concept of the present invention to those skilled in the art by referring to specific embodiments. The elements in the drawings are schematic and not drawn to scale. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0047] To enable those skilled in the art to better understand the solution of this application, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the scope of protection of this application.

[0048] In the following embodiments of this application, a virtual sensing method for the internal moisture content of a solid insulating material is taken as an example to elaborate on the solution of this application in detail. However, this embodiment does not limit the scope of protection of this application.

[0049] As Figure 1 shown, the present invention provides a virtual sensing method for the internal moisture content of a solid insulating material, and the method includes the following steps:

[0050] Step 11, obtain the characteristic parameters of the solid insulating material;

[0051] Step 12, construct a virtual detection environment model according to the characteristic parameters of the solid insulating material;

[0052] Step 13, apply a simulated electric field in the virtual detection environment model and obtain the change data of the polarization current and depolarization current;

[0053] Step 14, perform digital processing on the change data of the polarization current and depolarization current to generate corresponding spectrum data;

[0054] Step 15, perform numerical calculation according to the spectrum data and a preset objective function to obtain the prediction result of the internal moisture content of the solid insulating material;

[0055] Step 16, compare the prediction result with the actually detected moisture content data to obtain an evaluation result.

[0056] In the embodiments of the present invention, by accurately obtaining the characteristic parameters of the insulating material, the accuracy of the entire sensing process is ensured; the virtual detection environment model can be flexibly adjusted according to different insulating material characteristics, so as to adapt to the detection requirements of various materials and scenarios; by applying a simulated electric field and monitoring the current change, non-destructive detection of the insulating material is achieved, protecting the integrity of the material; digital processing can efficiently convert the current change data into spectrum data, facilitating subsequent analysis and calculation; by performing numerical calculation through a preset objective function, the internal moisture content of the insulating material can be predicted more accurately, providing strong support for preventive maintenance; by comparing with the actually detected data, the accuracy of the prediction result can be verified and the model can be calibrated as necessary, thereby continuously improving the reliability of the sensing method.

[0057] In a preferred embodiment of the present invention, obtaining characteristic parameters of solid insulating materials includes:

[0058] Obtaining original data on the characteristics of solid insulating materials;

[0059] Extracting key features related to the characteristics of solid insulating materials from the original data;

[0060] Clustering the key features, calculating the Euclidean distance of each data point to each cluster center, and assigning it to the corresponding cluster center; for each cluster, calculating the mean of all data points inside it and taking this mean as the new cluster center; repeating the steps until the preset number of iterations is reached. After clustering, different combinations of material characteristics will be obtained, and each category represents a group of materials with similar characteristics; the original data includes the resistivity, dielectric constant, breakdown voltage, and heat resistance of the material; the key features include the chemical composition, physical structure, and processing technology of the material.

[0061] In the embodiment of the present invention, through clustering analysis of the characteristic parameters of solid insulating materials, materials with similar characteristics can be quickly grouped into one category, and materials meeting specific performance requirements can be more efficiently screened out, thereby accelerating the product R & D and optimization process. Through clustering analysis, the performance characteristics of different material groups can be intuitively compared, so as to select the most suitable materials for specific application scenarios.

[0062] In another preferred embodiment of the present invention, for the above step 11, the characteristic parameters of solid insulating materials include: dielectric constant, dielectric loss factor, volume resistivity, dielectric strength, and insulation resistance.

[0063] In the embodiment of the present invention, through these key parameters, the electrical insulation performance of solid insulating materials can be accurately evaluated, providing reliable data support for the design and manufacture of electrical equipment; after clarifying the characteristic parameters, engineers can scientifically select suitable solid insulating materials according to specific application requirements to meet the specific insulation requirements of electrical equipment, which helps to prevent insulation failure and electrical faults during the design and use of electrical equipment, thereby improving the safety of the entire system.

[0064] In another preferred embodiment of the present invention, the above step 12 may include:

[0065] Collecting detailed data of solid insulating materials, which includes characteristic parameters such as its dielectric constant, dielectric loss factor, volume resistivity, dielectric strength, and insulation resistance; according to the actual shape and size of the solid insulating material, in computer-aided design software (such as 3D modeling software), starting from the collected size data, using line, surface, and solid tools to construct its accurate geometric model;

[0066] Create a new project in the simulation software, import the previously constructed 3D geometric model, and set the physical fields for the simulation as needed, such as electric field, temperature field, humidity field, etc.; in the simulation software, assign the characteristic parameters of the solid insulating material to the imported geometric model, such as dielectric constant, conductivity, thermal conductivity, etc., and ensure that these parameters match the actual material properties to obtain accurate simulation results.

[0067] According to the actual situation, set the boundary conditions for the model, such as electric potential, current, temperature, humidity, etc., apply a simulated electric field or other external loads to simulate the working conditions in the real environment. Mesh the model to generate a discretized mesh for simulation, and adjust the fineness of the mesh as needed to balance the calculation accuracy and calculation time. Set the solver parameters, such as the number of iterations, convergence criteria, etc. Run the simulation calculation, observe the behavior of the model in the virtual environment, analyze the simulation results, such as electric field distribution, current density, temperature change, etc., and adjust the model or simulation parameters as needed to improve the electrical or thermal performance of the design.

[0068] Input the characteristic parameters of the solid insulating material collected previously into the model of the virtual detection environment, and these parameters will affect the behavior of the model under the action of the electric field; in the virtual detection environment, set appropriate boundary conditions for the model, and these conditions include the distribution of the electric field, the flow direction of the current, the contact surface characteristics of the material, etc.

[0069] In the embodiment of the present invention, by constructing a geometric model that conforms to the actual shape and size and inputting accurate characteristic parameters, in the virtual detection environment, the shape, size, and characteristic parameters of the material can be adjusted. By simulating the performance under different conditions, the design of the material can be optimized. Constructing a virtual detection environment model can conduct simulation experiments on a computer, reducing the experimental cost and time. Conducting performance evaluation of the solid insulating material in the virtual environment can avoid safety problems caused by insufficient material performance in actual applications, thereby enhancing the overall safety of the electrical system.

[0070] In another preferred embodiment of the present invention, step 13 above may include:

[0071] According to the virtual detection environment model, apply a simulated electric field, and set the parameters of the electric field strength, direction, and action time according to the actual detection requirements to obtain the set simulated electric field;

[0072] Apply the electric field to the solid insulating material to generate a polarization current, and monitor the change of the polarization current in real time to obtain the change data of the polarization current and depolarization current.

[0073] In the embodiments of the present invention, by setting the intensity, direction, and action time of the electric field according to actual requirements, the electric field environment of the solid insulating material under actual working conditions can be accurately simulated, and electric field application and current monitoring can be carried out in a virtual environment without causing any damage to the solid insulating material, realizing non-destructive performance testing; the changes in polarization current and depolarization current can be monitored in real time, and detailed data can be obtained, facilitating subsequent data analysis and performance evaluation. The virtual detection environment model can significantly reduce the experimental cost and potential risks by obtaining the change data of polarization current and depolarization current.

[0074] In a preferred embodiment of the present invention, step 14 includes:

[0075] Step 141, periodically obtain electrical signal values according to the change data of polarization current and depolarization current, and convert the electrical signal values into a finite number of digital signal discrete values;

[0076] Step 142, reconstruct the polarization current and depolarization current data in the time domain according to the digital signal discrete values;

[0077] Step 143, calculate the corresponding spectrum data according to the polarization current and depolarization current data in the time domain.

[0078] In the embodiments of the present invention, by sampling at regular time intervals, the change data of polarization current and depolarization current are simplified into a series of discrete digital signal values, reducing the complexity of data processing. Digital signals have anti-interference ability compared with analog signals and can maintain the integrity of the signal during transmission; by reconstructing the polarization current and depolarization current data in the time domain, the waveform of the original signal can be restored, which helps to more accurately analyze the change characteristics of the current. The reconstructed time-domain data is a continuous and smooth curve, improving the observation and analysis accuracy of the current change details. The reconstructed time-domain data can intuitively show the change trends of polarization current and depolarization current; by calculating the spectrum data, the characteristics of polarization current and depolarization current in the frequency domain can be extracted. Spectrum analysis helps to identify specific frequency components related to the aging or failure of insulating materials, thereby realizing fault diagnosis and prevention. Combining the time-domain and frequency-domain data can comprehensively analyze the performance of insulating materials from multiple perspectives.

[0079] In the specific implementation process of the present invention, a time interval is determined. The time interval can capture the key changes in the polarization current and depolarization current. A high-speed data acquisition card is used to sample the two currents according to the time interval. Each sampling will obtain an electrical signal value representing the magnitude of the current at that time point. The continuous analog electrical signal values will be sent to an analog-to-digital converter to be converted into digital signals. The analog-to-digital converter will map each analog electrical signal value to a corresponding number, forming a series of discrete digital signal discrete values. According to the discrete digital signal values, using an interpolation algorithm, a continuous current waveform is estimated between the discrete data points. Through interpolation, a smooth current curve can be generated between the original sampling points, making the reconstructed time-domain current data closer to the actual current change situation. According to the reconstructed time-domain polarization current and depolarization current data, the corresponding spectral data is calculated through the fast Fourier transform algorithm, so as to analyze the amplitude and phase of different frequency components in the signal.

[0080] In a preferred embodiment of the present invention, the above step 142 may include:

[0081] By performing weighted processing on the digitized data to obtain the time-domain polarization current and depolarization current data;

[0082] where N is the window length covering the number of sample points, n is the sample point index in the window, a0 is the adjustment constant term coefficient, a1, a2, and a3 are the cosine coefficients, a4 is the linear term coefficient; w(n) is the value of the window function at the sample point n, and the samples are the data points collected from the polarization current and depolarization current signals.

[0083] In the embodiment of the present invention, the weighted processing can effectively smooth the data and reduce noise. By adjusting the window function and the window length, the local characteristics of the current signal can be highlighted. The weighted processing method provides high flexibility, and the result can be optimized by adjusting the window length, the type of window function, and the constant term, cosine coefficient, and linear term coefficients. By performing weighted processing on the data, the key characteristics of the polarization current and depolarization current can be captured more accurately. In the present invention, the window length defines the number of sample points used for weighted processing; the constant term coefficient is used to adjust the baseline value of the weighted function and can control the overall offset of the weighted function; the cosine coefficient can control the shape of the window function; the linear term coefficient allows the weighted function to have a certain slope; the window function determines the weight applied at each sample point and can optimize the spectral characteristics of the signal and reduce sidelobe leakage.

[0084] In a preferred embodiment of the present invention, the above step 143 may include:

[0085] According to the time-domain polarization current and depolarization current data, through Calculate the corresponding spectral data;

[0086] Where k represents the frequency, with a value range of 0 ≤ k ≤ N1 - 1, N1 represents the total number of sampling points, j represents the imaginary number, Δt represents the time difference between two adjacent sampling points, and f(n1) represents the value of the polarization current or depolarization current in the time domain at the n1th sampling point. is the complex exponential part of the transform, and n1 represents the index of the sampling point.

[0087] In the embodiments of the present invention, by calculating the spectral data, the performance of the polarization current and depolarization current at different frequencies can be directly observed; specific spectral features are associated with the aging, damage, or specific types of faults of the insulating material. By monitoring the changes in the spectral data, potential problems can be detected early, so as to take preventive measures to avoid equipment failures; in spectral analysis, noise and other interference signals can be identified, and storing spectral data may be more efficient than storing the original time-domain data.

[0088] In the present invention, the frequency represents the rate of periodic change of each component in the signal, the total number of sampling points represents the resolution and range of spectral analysis; the time difference between two adjacent sampling points is the reciprocal of the sampling rate, which determines the frequency resolution in spectral analysis. A smaller time difference means a higher sampling rate, so that higher frequency components can be resolved; the polarization current or depolarization current in the time domain at the nth sampling point represents the value of the polarization current or depolarization current collected at a specific time point.

[0089] In a preferred embodiment of the present invention, according to the change data of the polarization current and depolarization current, the electrical signal values are periodically obtained, and the electrical signal values are converted into a finite number of digital signal discrete values, including:

[0090] Determine the sampling frequency and set the sampling time length;

[0091] Sample the polarization current and depolarization current according to the set sampling frequency and time length;

[0092] Each sampling obtains an analog electrical signal value representing the current magnitude at the corresponding time point;

[0093] Send the analog electrical signal value into an analog-to-digital converter, and the analog-to-digital converter maps each analog electrical signal value to a corresponding number to obtain the digital signal discrete value.

[0094] In the present invention, by setting the sampling frequency and the sampling time length, the changes of the polarization current and the depolarization current can be accurately captured. The analog electrical signals are converted into discrete digital signal values, which facilitates efficient data processing, storage, and analysis using a computer. At the same time, the anti-interference ability and transmission efficiency of the data are improved. The processing of digital signals is more flexible, and operations such as filtering, transformation, and compression can be conveniently performed. Digital signals are easier to store for a long time and transmit remotely, without being restricted by time and space, which helps to achieve data sharing and remote monitoring.

[0095] In a preferred embodiment of the present invention, the analog-to-digital converter maps each analog electrical signal value to a corresponding digit, including:

[0096] Determine the range of the analog electrical signal and the range of the digital signal. Let the minimum value of the analog electrical signal be a min , and the maximum value be a max ;

[0097] According to the minimum value of the analog electrical signal being a min and the maximum value being a max , calculate the range width a r = a max - a min ;

[0098] Calculate the range width of the digital signal, and based on the range width of the digital signal, calculate the mapping ratio

[0099] For each analog electrical signal value a, perform the following steps to map it to the corresponding digital signal value:

[0100] Normalize the analog electrical signal value a within the range through n o = a - a min , where n o represents the normalized value;

[0101] Map the normalized value to the digital signal range using the scale factor through d v = n o × s f The digital value corresponding to the mapped value is represented by d v .

[0102] In the present invention, by determining the ranges of analog electrical signals and digital signals and calculating the mapping ratio, this method can achieve an accurate mapping from analog electrical signals to digital signals, ensuring that the digital signals can accurately reflect the characteristics of the original analog signals. By standardizing the values of analog electrical signals, this method eliminates the influence of dimensions and numerical magnitudes in the original data, enabling different analog electrical signals to be compared and processed on the same scale, thereby improving the efficiency and accuracy of data processing. Since the mapping ratio is dynamically calculated based on the ranges of analog electrical signals and digital signals, this method can adapt to analog electrical signals with different ranges and precisions, endowing the method with high flexibility and scalability. Converting analog signals to digital signals can greatly improve the efficiency of data processing. Digital signals are more easily stored, transmitted, and analyzed in a computer, and are less susceptible to noise and interference. Through the method of the present invention, a large number of analog electrical signals can be efficiently converted into digital signals, thus accelerating the speed of data processing.

[0103] In a preferred embodiment of the present invention, step 15 above includes:

[0104] Step 151, analyzing the spectral data to obtain the distribution of amplitude with frequency;

[0105] Step 152, defining an objective function for the relationship between spectral data and moisture content based on historical data;

[0106] Step 153, obtaining a prediction result of the moisture content inside the solid insulating material based on the spectral data and the objective function.

[0107] In the embodiment of the present invention, by analyzing the distribution of the amplitude of spectral data with frequency, the frequency composition of the signal can be understood; specific frequency components may indicate potential problems or faults in the equipment. Through spectral analysis, these problems can be detected early, which helps to take preventive measures in a timely manner; understanding the main frequency components in the signal helps to optimize the design and operating parameters of the system, thereby improving the overall performance of the system; by establishing an objective function based on historical data, a mathematical model capable of predicting the moisture content inside the solid insulating material can be constructed; based on the analysis of a large amount of historical data, the objective function becomes more accurate. Using historical data to define the objective function makes the decision-making process more scientific and data-driven, reducing the influence of subjective judgment; by predicting the moisture content inside the solid insulating material, preventive maintenance can be carried out in a timely manner, accurately predicting and promptly handling the moisture problem in the insulating material; through prediction and preventive maintenance, the downtime and maintenance costs caused by equipment failures can be reduced. Timely predicting and handling the moisture content problem inside the solid insulating material can reduce the risk of equipment failure, thereby improving the safety of equipment operation.

[0108] In the specific implementation process of the present invention, spectral data obtained through Fourier transform is acquired. The data usually includes amplitude and phase information corresponding to each frequency component. The amplitude information in the spectral data is extracted and matched with the corresponding frequency, and the data is organized into a list or array of frequency-amplitude pairs. A graph is used to display the distribution of amplitude with frequency. According to the analysis of the spectrogram, the amplitude change within the frequency range is determined, and a historical data set containing spectral data and corresponding moisture content is collected; according to the extracted spectral features, the features with the highest correlation with the moisture content are selected for modeling, a target function is constructed using statistical learning methods, and cross-validation is used for training and evaluation. The new spectral data is used as input and input into the trained target function. The processed spectral data is input into the target function for calculation to obtain the predicted moisture content result, and the prediction result is interpreted and analyzed to determine the moisture content level inside the solid insulating material.

[0109] In a preferred embodiment of the present invention, in step 152 above, the target function f(x i ,θ) has the following calculation formula:

[0110] f(x i ,θ) = θ0 + θ1log(x ik ) + θ2exp(-θ3x il );

[0111] Wherein, x ik and x il respectively represent the response values of the i-th sample at frequencies k and l; θ0, θ1, θ2, θ3 are parameter vectors; θ is the parameter vector containing all parameters.

[0112] In the embodiment of the present invention, the target function allows considering the response values at multiple frequencies simultaneously. The responses at different frequencies reflect different physical processes or states. Considering the responses comprehensively can provide accurate prediction and analysis. By adjusting the parameters θ0, θ1, θ2, θ3, different application scenarios and data sets can be adapted; the linear form of the function makes the result have an intuitive interpretability; according to the linear property of the function, standard optimization algorithms can be used to find the optimal parameter values.

[0113] In a preferred embodiment of the present invention, step 153 above may include:

[0114] According to the spectral data and the target function, through calculation is performed to obtain the predicted result of the moisture content inside the solid insulating material;

[0115] Wherein, y i is the actual moisture content of the i-th sample, and x iis the spectral data of the i-th sample, f(x i , θ) is the moisture content predicted based on the spectral data x i and the parameter vector θ. θ is the parameter vector, m is the number of parameters, λ is the regularization coefficient, |θ j | is the absolute value of the parameter vector θ, γ is another regularization coefficient, n2 is the number of samples, m is the number of parameters, p is the number of regularization terms, min represents finding the minimum value to find the parameter vector θ that minimizes the objective function value; X represents the objective function value corresponding to the optimal solution of the parameter vector θ obtained by minimizing the expression within the parentheses.

[0116] In the embodiments of the present invention, by minimizing the difference between the actual moisture content and the predicted moisture content, the optimal parameters can be found to improve the prediction accuracy. The introduction of the regularization term helps prevent the model from overfitting; by comprehensively considering the difference between the actual moisture content and the predicted moisture content and the regularization term, the model parameters that perform well on the training data and also have good generalization ability on unseen data can be found; the formula allows adjusting the regularization coefficient and other regularization terms according to the application scenario, making the formula flexible and scalable, and can adapt to different prediction problems and data sets; the formula takes into account the complexity and generalization ability of the model, provides a complete prediction framework, and can select appropriate regularization terms and constraints according to specific problems and data sets for effective prediction and analysis.

[0117] As Figure 2 shown, a control system 20 for a virtual sensing method of the moisture content inside a solid insulating material includes the following steps:

[0118] An acquisition module 21, configured to acquire the characteristic parameters of the solid insulating material; construct a virtual detection environment model according to the characteristic parameters of the solid insulating material; apply a simulated electric field in the virtual detection environment model, and acquire the change data of the polarization current and the depolarization current.

[0119] A processing module 22, configured to digitally process the change data of the polarization current and the depolarization current to generate corresponding spectral data; perform numerical calculations according to the spectral data and a preset objective function to obtain the prediction result of the moisture content inside the solid insulating material; compare the prediction result with the actually detected moisture content data to obtain an evaluation result.

[0120] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A virtual sensing method for the moisture content inside a solid insulating material, characterized in that, Including the following steps: Obtain the characteristic parameters of the solid insulation material; Construct a virtual detection environment model according to the characteristic parameters of the solid insulation material; Apply a simulated electric field in the virtual detection environment model and obtain the change data of the polarization current and depolarization current; Digitally process the change data of the polarization current and depolarization current to generate corresponding spectrum data; wherein, according to the change data of the polarization current and depolarization current, periodically obtain the electrical signal values, convert the electrical signal values into a finite number of digital signal discrete values; determine the sampling frequency and set the sampling time length; sample the polarization current and depolarization current according to the set sampling frequency and time length; each sampling obtains an analog electrical signal value representing the current magnitude at the corresponding time point; send the analog electrical signal value into an analog-to-digital converter, and the analog-to-digital converter maps each analog electrical signal value to a corresponding number to obtain the digital signal discrete value; Reconstruct the polarization current and depolarization current data in the time domain according to the digital signal discrete values; calculate the corresponding spectrum data according to the polarization current and depolarization current data in the time domain; Perform numerical calculations according to the spectrum data and a preset objective function to obtain the prediction result of the moisture content inside the solid insulation material; wherein, analyze the spectrum data to obtain the distribution of the amplitude with frequency; define the objective function of the relationship between the spectrum data and the moisture content according to the historical data; obtain the prediction result of the moisture content inside the solid insulation material according to the spectrum data and the objective function; Compare the prediction result with the moisture content data detected actually to obtain the evaluation result.

2. The virtual sensing method for the internal moisture content of the solid insulating material according to claim 1, wherein, Obtain the characteristic parameters of the solid insulation material, including: Obtain the original data regarding the characteristics of the solid insulation material; Extract the key features related to the characteristics of the solid insulation material from the original data; Cluster the key features, calculate the distance from each data point to each cluster center, and assign it to the corresponding cluster center; for each cluster, calculate the mean value of all data points inside it and use this mean value as the new cluster center; repeat the steps until the preset number of iterations is reached. After clustering, different material characteristic combinations will be obtained, and each category represents a group of materials with similar characteristics.

3. The virtual sensing method for the internal moisture content of the solid insulating material according to claim 2, characterized in that, The original data includes the resistivity, dielectric constant, breakdown voltage, and heat resistance of the material; The key features include the chemical composition, physical structure, and processing technology of the material.

4. The virtual sensing method for the internal moisture content of the solid insulating material according to claim 1, characterized in that, The analog-to-digital converter maps each analog electrical signal value to a corresponding number, including: Determine the range of the analog electrical signal and the range of the digital signal. Let the minimum value of the analog electrical signal be a min , and the maximum value be a max ; According to the minimum value of the analog electrical signal being a min and the maximum value being a max , calculate the range width a of the analog signal r = a max - a min ; Calculate the range width of the digital signal, and calculate the mapping ratio according to the range width of the digital signal For each analog electrical signal value a, perform the following steps to map it to the corresponding digital signal value: Normalize the analog electrical signal value a to within a range through n o = a - a min where n o represents the normalized value; Use a scale factor to map the normalized value to the digital signal range through d v = n o × s f where d v represents the corresponding digital value after mapping.

5. A virtual sensing system for the moisture content inside a solid insulating material, which system implements the method according to any one of claims 1-4, characterized in that: Including: An acquisition module for obtaining the characteristic parameters of the solid insulation material; Construct a virtual detection environment model according to the characteristic parameters of the solid insulation material; Apply a simulated electric field in the virtual detection environment model and obtain the change data of the polarization current and depolarization current; A processing module for digitally processing the change data of the polarization current and the depolarization current to generate corresponding spectrum data; performing numerical calculations based on the spectrum data and a preset objective function to obtain a prediction result of the moisture content inside the solid insulating material; comparing the prediction result with the actually detected moisture content data to obtain an evaluation result.

6. A computing device, characterized in that, Comprising: One or more processors; A storage device for storing one or more programs, which when executed by the one or more processors cause the one or more processors to implement the method according to any one of claims 1-4.

7. A computer-readable storage medium, characterized in that, A program is stored in the computer-readable storage medium, and when the program is executed by a processor, the method according to any one of claims 1-4 is implemented.

Images