Virtual sensing method and system for internal moisture content of solid insulating material

Abstract

The 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 insulating material, and the method comprises the following steps: obtaining the characteristic parameters of the solid insulating material; constructing a virtual detection environment model according to the characteristic parameters of the solid insulating material; applying a simulation electric field in the virtual detection environment model, and obtaining change data of polarization current and depolarization current; performing digital processing on the change data of the polarization current and the depolarization current to generate corresponding frequency spectrum data; performing numerical calculation according to the frequency spectrum data and a preset objective function to obtain a prediction result of the internal moisture content of the solid insulating material; comparing the prediction result with the actually detected moisture content data to obtain an evaluation result; according to the method, prediction can be carried out through the virtual detection environment model, physical damage to the solid insulating material is not needed, and the integrity and reusability of the material are guaranteed.

Description

Technical Field

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

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

[0003] Some traditional moisture content detection methods require physical sampling and destructive testing of insulating materials, which is not only time-consuming and labor-intensive, but also unable to monitor the moisture changes inside the material in real time. In addition, these methods may not be able to fully reflect the moisture distribution inside 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 shortcomings of the prior art and provide a virtual sensing method and system for the internal moisture content of solid insulating materials, which can be predicted through a virtual detection environment model without physically destroying the solid insulating materials, thereby ensuring the integrity and reusability of the materials.

[0005] In order to solve the above technical problems, the basic concept of the technical solution adopted by the present invention is:

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

[0007] Obtaining characteristic parameters of solid insulation materials;

[0008] Construct a virtual testing environment model based on the characteristic parameters of solid insulation materials;

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

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

[0011] Numerical calculation is performed based on the spectrum data and the preset objective function to obtain the prediction result of the moisture content inside the solid insulating material;

[0012] The predicted results are compared with the actual detected moisture content data to obtain the evaluation results.

[0013] Furthermore, the characteristic parameters of the solid insulation material are obtained, including:

[0014] Obtain raw data on the properties of solid insulating materials;

[0015] Extract key features related to the properties of solid insulation materials from raw data;

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

[0017] Furthermore, the raw 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] Furthermore, the change data of the polarization current and the depolarization current are digitally processed to generate corresponding spectrum data, including:

[0019] According to the change data of polarization current and depolarization current, the electrical signal value is periodically acquired, and the electrical signal value is converted into a limited number of digital signal discrete values;

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

[0021] The corresponding spectrum data is calculated based on the polarization current and depolarization current data in the time domain.

[0022] Furthermore, according to the change data of the polarization current and the depolarization current, the electrical signal value is periodically acquired, and the electrical signal value is converted into a limited number of digital signal discrete values, including:

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

[0024] The polarization current and depolarization current are sampled 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] The analog electrical signal value is sent to the analog-to-digital converter, and the analog-to-digital converter maps each analog electrical signal value to a corresponding number to obtain a digital signal discrete value.

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

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

[0029] According to the minimum value of the analog electrical signal is a min The maximum value is 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 calculate the mapping ratio based on the range width of the digital signal

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

[0032] The analog electrical signal value a is passed through n o =a-a min Normalized to the range, where n o represents a standardized value;

[0033] Use the scaling factor d v =n o ×s f Map the normalized value to the digital signal range, d v Indicates the corresponding digital value after mapping.

[0034] Furthermore, numerical calculations are performed based on the spectrum data and the preset objective function to obtain the prediction results of the moisture content inside the solid insulating material, including:

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

[0036] Based on historical data, define the objective function of the relationship between spectrum data and moisture content;

[0037] According to the spectrum data and the objective function, the prediction result of the moisture content inside the solid insulating material is obtained.

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

[0039] The acquisition module is used 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.

[0040] The processing module is used to digitally process the change data of the polarization current and the depolarization current to generate corresponding spectrum data; perform 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; and compare the prediction result with the actual detected moisture content data to obtain an evaluation result.

[0041] After adopting the above technical scheme, the present invention has the following beneficial effects compared with the prior art: by making predictions through a virtual detection environment model, there is no need to physically destroy the solid insulating material, thereby 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; by performing numerical calculations using a preset objective function, the moisture content inside the solid insulating material can be predicted, which helps to promptly discover potential moisture problems and take corresponding preventive measures; the virtual perception method reduces the frequency and need for physical detection, thereby reducing maintenance costs and labor costs, and can predict and deal with potential problems in advance; through accurate detection of the moisture content inside the insulating material, potential safety hazards can be discovered and dealt with in a timely manner, thereby improving the overall safety of the electrical system.

[0042] The specific implementation modes of the present invention are further described in detail below in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The 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 of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. Some specific embodiments of the present application will be described in detail in an illustrative and non-restrictive manner with reference to the drawings. The same reference numerals in the drawings indicate the same or similar components or parts. It should be understood by those skilled in the art that these drawings are not necessarily drawn to scale. In the drawings:

[0044] Figure 1 It is a flow chart of a virtual sensing method for the internal moisture content of a solid insulating material according to the present invention.

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

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

[0047] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work should fall within the scope of protection of the present application.

[0048] The following embodiment of the present application takes a virtual sensing method of the internal moisture content of a solid insulating material as an example to explain in detail the scheme of the present application, but this embodiment cannot limit the protection scope of the present application.

[0049] like Figure 1 As shown, the present invention provides a virtual sensing method for the moisture content inside a solid insulating material, the method comprising the following steps:

[0050] Step 11, obtaining characteristic parameters of solid insulating materials;

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

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

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

[0054] Step 15, 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;

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

[0056] In the embodiments of the present invention, the accuracy of the entire perception process is ensured by accurately acquiring the characteristic parameters of the insulating material; the virtual detection environment model can be flexibly adjusted according to the characteristics of different insulating materials, thereby adapting to the detection needs 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, thereby protecting the integrity of the material; digital processing can efficiently convert the current change data into spectrum data, which is convenient for subsequent analysis and calculation; numerical calculation through a preset objective function can more accurately predict the moisture content inside the insulating material, providing strong support for preventive maintenance; by comparing with the actual detection data, the accuracy of the prediction results can be verified, and the model can be calibrated as necessary, thereby continuously improving the reliability of the perception method.

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

[0058] Obtain raw data on the properties of solid insulating materials;

[0059] Extract key features related to the properties of solid insulation materials from raw data;

[0060] Cluster the key features, calculate the Euclidean distance from each data point to each cluster center, and assign them to the corresponding cluster center; for each cluster, calculate the mean of all data points within it, and use the mean as the new cluster center; repeat the steps until the preset number of iterations is reached. After clustering is completed, different combinations of material properties will be obtained, and each category represents a group of materials with similar properties; 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, by clustering the characteristic parameters of solid insulating materials, materials with similar characteristics can be quickly classified into one category, and materials that meet specific performance requirements can be screened out more efficiently, thereby accelerating the product development and optimization process. Through cluster analysis, the performance characteristics of different material groups can be intuitively compared, so as to select the most suitable material for a specific application scenario.

[0062] In another preferred embodiment of the present invention, in the above step 11, the characteristic parameters of the solid insulating material 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] Collect detailed data of solid insulating materials, including characteristic parameters such as dielectric constant, dielectric loss factor, volume resistivity, dielectric strength and insulation resistance; Based on the actual shape and size of the solid insulating materials, use line, surface and solid tools in computer-aided design software (such as 3D modeling software) to start building its precise geometric model according to the collected dimensional data;

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

[0067] According to the actual situation, set boundary conditions for the model, such as electric potential, current, temperature, humidity, etc., and apply simulated electric fields or other external loads to simulate the working conditions in the real environment. Mesh the model to generate discretized meshes for simulation, adjust the fineness of the mesh as needed to balance the calculation accuracy and calculation time, and set solver parameters such as the number of iterations and convergence criteria. 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] The characteristic parameters of solid insulating materials collected previously are input into the model of the virtual detection environment. These parameters will affect the behavior of the model under the action of the electric field. In the virtual detection environment, appropriate boundary conditions are set for the model. These conditions include the distribution of the electric field, the direction of the current, the contact surface characteristics of the material, etc.

[0069] In an embodiment of the present invention, by constructing a geometric model that matches the actual shape and size and inputting accurate characteristic parameters, the shape, size and characteristic parameters of the material can be adjusted in a virtual detection environment. By simulating the performance under different conditions, the design of the material can be optimized. By constructing a virtual detection environment model, simulation experiments can be carried out on a computer to reduce experimental costs and time. Performance evaluation of solid insulating materials in a 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, the above step 13 may include:

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

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

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

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

[0075] Step 141, periodically acquiring electrical signal values ​​according to the change data of the polarization current and the depolarization current, and converting the electrical signal values ​​into a limited number of digital signal discrete values;

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

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

[0078] In an embodiment of the present invention, by sampling at regular time intervals, the change data of the polarization current and the depolarization current are simplified into a series of discrete digital signal values, thereby reducing the complexity of data processing. Compared with analog signals, digital signals have anti-interference capabilities and can maintain signal integrity 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, which improves the observation and analysis accuracy of the current change details. The reconstructed time domain data can intuitively show the change trend of the polarization current and the depolarization current. By calculating the spectral data, the characteristics of the polarization current and the depolarization current in the frequency domain can be extracted. Spectral analysis helps to identify specific frequency components related to aging or failure of insulating materials, thereby realizing fault diagnosis and prevention. Combining the time domain and frequency domain data, the performance of the insulating material can be comprehensively analyzed from multiple angles.

[0079] In the specific implementation process of the present invention, a time interval is determined, and the time interval can capture the key changes of the polarization current and the depolarization current. A high-speed data acquisition card is used to sample the two currents according to the time interval, and each sampling will obtain an electrical signal value representing the current size at that time point; the continuous analog electrical signal values ​​will be sent to the analog-to-digital converter to be converted into digital signals, and the analog-to-digital converter will map each analog electrical signal value to a corresponding number to form a series of discrete digital signal discrete values; according to the digital signal discrete values, an interpolation algorithm is used to estimate a continuous current waveform between discrete data points, and through interpolation, a smooth current curve can be generated between the original sampling points, so that the reconstructed time-domain current data will be closer to the actual current change; according to the reconstructed time-domain polarization current and depolarization current data, the corresponding spectrum data is calculated by 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] pass Performing weighted processing on the digitally processed data to obtain polarization current and depolarization current data in the time domain;

[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 cosine coefficients, and a4 is the linear term coefficient; w(n) is the value of the window function at sample point n, and the sample is the data point collected from the polarization current and depolarization current signals.

[0083] In the embodiment of the present invention, weighted processing can effectively smooth data and reduce noise. By adjusting the window function and window length, the local characteristics of the current signal can be highlighted. The weighted processing method provides high flexibility, and the results can be optimized by adjusting the window length, window function type, constant term, cosine coefficient and linear term coefficient. By weighted processing the data, the key characteristics of 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, which 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, which 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 polarization current and depolarization current data in the time domain, Calculate the corresponding spectrum data;

[0086] Where k represents the frequency, ranging from 0≤k≤N1-1, N1 represents the total number of sampling points, j represents an 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 transformation, and n1 represents the index of the sampling point.

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

[0088] In the present invention, frequency represents the rate of periodic change of each component in the signal, and the total number of sampling points represents the resolution and range of spectrum analysis; the time difference between two adjacent sampling points is the inverse of the sampling rate, which determines the frequency resolution in the spectrum analysis, and a smaller time difference means a higher sampling rate, thereby being able to distinguish higher frequency components; the polarization current or depolarization current in the time domain represents the value of the polarization current or depolarization current collected at a specific time point at the nth sampling point.

[0089] In a preferred embodiment of the present invention, based on the change data of the polarization current and the depolarization current, the electrical signal value is periodically acquired, and the electrical signal value is converted into a limited number of digital signal discrete values, including:

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

[0091] The polarization current and depolarization current are sampled 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] The analog electrical signal value is sent to the analog-to-digital converter, and the analog-to-digital converter maps each analog electrical signal value to a corresponding number to obtain a digital signal discrete value.

[0094] In the present invention, by setting the sampling frequency and the sampling time length, the changes in polarization current and depolarization current can be accurately captured. The analog electrical signal is converted into a digital signal discrete value, which is convenient for efficient data processing, storage and analysis using a computer, while improving the anti-interference ability and transmission efficiency of the data. The processing method 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, are not limited by time and space, and are conducive to 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 digital number, including:

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

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

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

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

[0100] The analog electrical signal value a is passed through n o =a-a min Normalized to the range, where n o represents a standardized value;

[0101] Use the scaling factor d v =n o ×s f Map the normalized value to the digital signal range, d v Indicates the corresponding digital value after mapping.

[0102] In the present invention, by determining the range of analog electrical signals and digital signals and calculating the mapping ratio, the method can realize accurate mapping of analog electrical signals to digital signals, ensuring that the digital signals can accurately reflect the characteristics of the original analog signals. By standardizing the analog electrical signal values, the method eliminates the influence of the dimension and numerical value in the original data, so that different analog electrical signals can be compared and processed on the same scale, improving the efficiency and accuracy of data processing. Since the mapping ratio is dynamically calculated based on the range of analog electrical signals and digital signals, the method can adapt to analog electrical signals of different ranges and precisions, making the method highly flexible and scalable. Converting analog signals to digital signals can greatly improve the efficiency of data processing. Digital signals are easier to store, transmit and analyze in computers, and are not easily affected by noise and interference. Through the method of the present invention, a large number of analog electrical signals can be efficiently converted into digital signals, thereby speeding up data processing.

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

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

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

[0106] Step 153, obtaining a prediction result of the moisture content inside the solid insulating material according to the frequency spectrum data and the objective function.

[0107] In an embodiment of the present invention, by analyzing the distribution of the amplitude of the spectrum data with frequency, the frequency composition of the signal can be understood; specific frequency components may indicate potential problems or failures in the equipment. Through spectrum analysis, these problems can be discovered early, which helps to take preventive measures in time; 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; establishing an objective function through historical data can construct a mathematical model that can predict the moisture content inside the solid insulating material; based on the analysis of a large amount of historical data, the objective function is made more accurate, and 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 time, and moisture problems in the insulating material can be accurately predicted and dealt with in time; through predictive and preventive maintenance, downtime and maintenance costs caused by equipment failure can be reduced, and timely prediction and handling of moisture content problems inside solid insulating materials can reduce the risk of equipment failure, thereby improving the safety of equipment operation.

[0108] In the specific implementation process of the present invention, spectrum data obtained by Fourier transform calculation is obtained, and the data generally includes amplitude and phase information corresponding to each frequency component. The amplitude information in the spectrum data is extracted and matched with the corresponding frequency. The data is organized into a list or array of frequency-amplitude pairs, and a chart is used to display the distribution of amplitude with frequency. According to the analysis of the spectrum graph, the amplitude change within the frequency range is determined, and a historical data set containing spectrum data and corresponding moisture content is collected; according to the extracted spectrum features, the features with the highest correlation with moisture content are selected for modeling, and a statistical learning method is used to construct an objective function. Cross-validation is used for training and evaluation, and new spectrum data is used as input and input into the trained objective function. The processed spectrum data is input into the objective function, and calculation is performed to obtain a predicted moisture content result, and the predicted 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 the above step 152, the objective function f(x i ,θ) is calculated as:

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

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

[0112] In an embodiment of the present invention, the objective function allows the response values ​​at multiple frequencies to be considered simultaneously. The responses at different frequencies reflect different physical processes or states. Comprehensive consideration of the responses can provide accurate prediction and analysis. By adjusting the parameters θ0, θ1, θ2, and θ3, it can adapt to different application scenarios and data sets. The linear form of the function makes the result intuitively interpretable. Based on the linear properties of the function, the optimal parameter value can be found using a standard optimization algorithm.

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

[0114] According to the spectrum data and the objective function, Calculations are performed to obtain predictions of the moisture content inside solid insulating materials;

[0115] Among them, y i is the actual moisture content of the ith sample, x iis the spectrum data of the i-th sample, f(x i ,θ) is based on the spectrum data x i and the moisture content predicted by 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 means 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 in the brackets.

[0116] In an embodiment 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 accuracy of the prediction, and the introduction of the regularization term helps to prevent the model from overfitting; by comprehensively considering the difference between the actual moisture content and the predicted moisture content and the regularization term, model parameters that perform well on training data and have good generalization ability on unseen data can be found; the formula allows the regularization coefficient and other regularization terms to be adjusted according to the application scenario, so that the formula is 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, and provides a complete prediction framework, which can select appropriate regularization terms and constraints according to specific problems and data sets to perform effective prediction and analysis.

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

[0118] The acquisition module 21 is used 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] The processing module 22 is used to digitally process the change data of the polarization current and the depolarization current to generate corresponding spectrum data; perform 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; and compare the prediction result with the actual 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 it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, 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: The following steps are involved: Obtain characteristic parameters of solid insulation materials; Construct a virtual testing environment model based on the characteristic parameters of solid insulation materials; Applying a simulated electric field in the virtual detection environment model and obtaining the change data of polarization current and depolarization current; Digitally processing the change data of polarization current and depolarization current to generate corresponding spectrum data; Numerical calculation is performed based on the spectrum data and the preset objective function to obtain the prediction result of the moisture content inside the solid insulating material; The predicted results are compared with the actual detected moisture content data to obtain the evaluation results.

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

3. The virtual sensing method for the moisture content inside a solid insulating material according to claim 2, characterized in that: The original data include the material’s resistivity, dielectric constant, breakdown voltage, and heat resistance; Key characteristics include the material's chemical composition, physical structure and processing.

4. The virtual sensing method for the moisture content inside a solid insulating material according to claim 3 is characterized in that: The change data of polarization current and depolarization current are digitized to generate corresponding spectrum data, including: According to the change data of polarization current and depolarization current, the electrical signal value is periodically acquired, and the electrical signal value is converted into a limited number of digital signal discrete values; Reconstruct the polarization current and depolarization current data in the time domain according to the discrete value of the digital signal; The corresponding spectrum data is calculated based on the polarization current and depolarization current data in the time domain.

5. The virtual sensing method for the moisture content inside a solid insulating material according to claim 4, characterized in that: According to the change data of polarization current and depolarization current, the electrical signal value is periodically obtained and converted into a limited number of digital signal discrete values, including: Determine the sampling frequency and set the sampling time length; The polarization current and depolarization current are sampled 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; The analog electrical signal value is sent to the analog-to-digital converter, and the analog-to-digital converter maps each analog electrical signal value to a corresponding number to obtain a digital signal discrete value.

6. The virtual sensing method for the moisture content inside a solid insulating material according to claim 5, characterized in that: The analog-to-digital converter maps each analog electrical signal value to a corresponding digital value, including: Determine the range of analog electrical signals and the range of digital signals. Let the minimum value of the analog electrical signal be a. min , the maximum value is a max ; According to the minimum value of the analog electrical signal is a min The maximum value is 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 based on 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: The analog electrical signal value a is passed through n o =a-a min Normalized to the range, where n o represents a standardized value; Use the scaling factor d v =n o ×s f Map the normalized value to the digital signal range, d v Indicates the corresponding digital value after mapping.

7. The virtual sensing method for the moisture content inside a solid insulating material according to claim 6, characterized in that: Numerical calculations are performed based on the spectrum data and the preset objective function to obtain the prediction results of the moisture content inside the solid insulation material, including: Analyze spectral data to obtain the distribution of amplitude over frequency; Based on historical data, define the objective function of the relationship between spectrum data and moisture content; According to the spectrum data and the objective function, the prediction result of the moisture content inside the solid insulating material is obtained.

8. A virtual sensing system for the moisture content inside a solid insulating material, the system implementing the method according to any one of claims 1 to 7, characterized in that: include: An acquisition module, used for acquiring characteristic parameters of solid insulating materials; Construct a virtual testing environment model based on the characteristic parameters of solid insulation materials; Applying a simulated electric field in the virtual detection environment model and obtaining the change data of polarization current and depolarization current; The processing module is used to digitally process the change data of the polarization current and the depolarization current to generate corresponding spectrum data; perform 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; and compare the prediction result with the actual detected moisture content data to obtain an evaluation result.

9. A computing device, characterized in that include: one or more processors; A storage device, used for storing one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a program, which, when executed by a processor, implements the method according to any one of claims 1 to 7.

Images