Insulator PRTV coating thickness detection method

Abstract

The invention provides a method for detecting the thickness of a PRTV coating of an insulator, and the method comprises the steps: measuring the amplitude and phase data of the PRTV coating of the insulator through an amplitude and phase method, carrying out the preprocessing of the measured amplitude and phase data, building a relation image curve of the thicknesses of different positions of the PRTV coating and the amplitude and phase parameters according to the corresponding thickness of the insulator, and carrying out the detection of the thickness of the PRTV coating. The method comprises the following steps: constructing an insulator PRTV coating thickness detection model, taking a PRTV coating thickness and amplitude phase parameter relation image curve based on different position thicknesses of a PRTV coating as a training set, training the model based on the training set, evaluating model performance through verification data independent of the training set, adjusting hyper-parameters, optimizing the model, and deploying the optimized model. And carrying out thickness detection on the to-be-detected insulator PRTV coating. The method does not need to carry out a complex real-model test, saves the test and labor cost, is simple and convenient, can be applied to different operation and maintenance conditions, and has good expandability.

Description

Technical Field

[0001] The invention relates to the technical field of power transmission lines, and in particular to a method for detecting the thickness of an insulator PRTV coating. Background Art

[0002] As an important component in the power system, the performance of transmission line insulators directly affects the safe operation of the system. The surface hydrophobicity and pollution flashover resistance of a large number of ceramic suspension insulators in existing transmission lines are poor. In order to enhance the pollution flashover resistance of insulators in harsh environments, a layer of PRTV material is usually coated on its surface. The PRTV coating can improve the hydrophobicity of the insulator, reduce surface pollution, and effectively prevent pollution flashover accidents. At present, the on-site coating construction of the PRTV coating on the surface of the insulator is mainly spraying and brushing, which is easily affected by human and environmental factors. Existing research and operating experience have shown that the quality of PRTV coating is poor, such as damage, impurities, debonding, uneven thickness or non-compliance with standards; at the same time, under complex and changeable operating conditions such as heavy rain, ice, fog, and low pressure, PRTV coatings will still show certain aging phenomena, which will seriously affect the external insulation performance and pose a hidden danger to the safety and reliability of the equipment. At the same time, with the passage of time, the PRTV coating will gradually age, wear or fail due to environmental factors (such as ultraviolet radiation, acid rain, wind and sand, etc.). The thickness of the coating directly affects its functional effect. Therefore, it is very important to detect the PRTV thickness, surface condition and internal defects of the suspension insulator surface before the sprayed PRTV coating leaves the factory and after a certain number of years of operation.

[0003] Traditional coating thickness detection methods include manual measurement or offline testing through random sampling, but these methods have the following significant drawbacks:

[0004] Low detection efficiency. Manual detection can usually only cover local areas and cannot perform rapid and comprehensive detection of large-scale equipment.

[0005] Poor accuracy. Manual testing relies on the operator’s experience and is greatly affected by human factors, making it difficult to ensure the accuracy and consistency of test data.

[0006] The existing detection methods have poor adaptability and are difficult to be flexibly applied to insulators of different types and in different environments, especially for high-altitude insulators in harsh outdoor environments, where detection is more difficult.

[0007] Therefore, it is necessary to design a method for detecting the thickness of the insulator PRTV coating. Summary of the invention

[0008] In order to overcome the deficiencies of the prior art, an object of the present invention is to provide a method for detecting the thickness of an insulator PRTV coating.

[0009] To achieve the above object, the present invention provides the following solutions:

[0010] The present invention provides a method for detecting the thickness of an insulator PRTV coating, comprising:

[0011] The amplitude and phase data of the insulator PRTV coating are measured by the amplitude phase method;

[0012] The measured amplitude and phase data are preprocessed, and the image curve of the relationship between the thickness of the PRTV coating at different positions and the amplitude and phase parameters is established according to the corresponding insulator thickness;

[0013] The insulator PRTV coating thickness detection model is constructed, and the image curves based on the relationship between the thickness of the PRTV coating at different positions and the amplitude phase parameters are used as the training set, and the model is trained based on the training set;

[0014] Evaluate model performance and adjust hyperparameters using validation data independent of the training set;

[0015] Optimize the model;

[0016] The optimized model is deployed to detect the thickness of the PRTV coating on the insulator to be tested.

[0017] Preferably, the amplitude and phase data of the insulator PRTV coating are measured by an amplitude phase method, specifically:

[0018] Perform amplitude measurement and obtain amplitude data;

[0019] Perform phase measurement and obtain phase data.

[0020] Preferably, amplitude measurement is performed to obtain amplitude data, specifically:

[0021] According to the characteristics of the coating to be measured and the measurement requirements, select an amplitude measuring instrument with appropriate frequency range and measurement accuracy;

[0022] Calibrate the selected amplitude measuring instrument;

[0023] According to the material and thickness of the coating to be measured, select a suitable amplitude measurement probe and install it on the amplitude measurement instrument;

[0024] Clean the sample to be tested;

[0025] Set the parameters of the amplitude measuring instrument according to the material and thickness of the coating to be measured;

[0026] Place the amplitude measurement probe on the surface of the sample to be measured, ensure that the probe is in close contact with the sample to be measured, and adjust the position and direction of the probe to obtain the best measurement effect;

[0027] Start the amplitude measuring instrument to measure the amplitude data of the coating under test;

[0028] Filtering and smoothing the measured amplitude data;

[0029] constructing an amplitude spectrum based on the processed amplitude data;

[0030] According to the changing trend and characteristics of the amplitude spectrum, combined with the material and characteristics of the measured coating, the thickness of the measured coating can be preliminarily inferred.

[0031] Preferably, phase measurement is performed to obtain phase data, specifically:

[0032] According to the characteristics of the coating to be measured and the measurement requirements, select a phase measurement instrument with appropriate frequency range and measurement accuracy;

[0033] Calibrate selected phase measurement instruments;

[0034] According to the material and thickness of the coating to be measured, select a suitable phase measurement probe and install it on the phase measurement instrument;

[0035] Clean the sample to be tested;

[0036] Set the parameters of the phase measuring instrument according to the material and thickness of the coating to be measured;

[0037] Place the phase measurement probe on the surface of the sample to be measured, ensure that the probe is in close contact with the sample to be measured, and adjust the position and direction of the probe to obtain the best measurement effect;

[0038] Start the phase measuring instrument to measure the phase data of the measured coating;

[0039] Filtering and smoothing the measured phase data;

[0040] constructing a phase spectrum based on the processed phase data;

[0041] By comparing the measured phase data with the reference phase information, the phase difference is calculated;

[0042] Based on the phase difference, combined with the material and characteristics of the measured coating, the thickness of the measured coating can be preliminarily inferred.

[0043] Preferably, the measured amplitude and phase data are preprocessed, and an image curve of the relationship between the thickness of the PRTV coating at different positions and the amplitude and phase parameters is established according to the corresponding insulator thickness, specifically:

[0044] The measured amplitude data and the thickness of the measured coating inferred from the amplitude data, the measured phase data and the thickness of the measured coating inferred from the phase data are obtained, and data cleaning and standardization or normalization processing are performed on them. Based on the processed data, an image curve of the relationship between the thickness of the PRTV coating at different positions and the amplitude and phase parameters is generated.

[0045] Preferably, an insulator PRTV coating thickness detection model is constructed, and an image curve based on the relationship between the thickness at different positions of the PRTV coating and the amplitude phase parameter is used as a training set, and the model is trained based on the training set, specifically:

[0046] Define layers and nodes, stack various types of layers to form a deep network, build an insulator PRTV coating thickness detection model based on the deep network, and use the image curve of the relationship between the thickness and amplitude phase parameters of the PRTV coating at different positions as the training set, and train the model based on the training set.

[0047] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0048] The present invention provides a method for detecting the thickness of an insulator PRTV coating, the method comprising measuring the amplitude and phase data of the insulator PRTV coating by an amplitude-phase method, preprocessing the measured amplitude and phase data, and establishing an image curve of the relationship between the thickness of the PRTV coating at different positions and the amplitude-phase parameters according to the corresponding insulator thickness, constructing an insulator PRTV coating thickness detection model, and using the image curve of the relationship between the thickness of the PRTV coating at different positions and the amplitude-phase parameters as a training set, training the model based on the training set, evaluating the model performance through verification data independent of the training set, adjusting hyperparameters, optimizing the model, deploying the optimized model, and performing thickness detection on the insulator PRTV coating to be tested. The present invention can quickly detect the thickness of the PRTV coating along the surface of power equipment under different operating conditions through deep learning, without the need to carry out complex true-type tests, greatly saving test and labor costs, being simple and convenient, and being able to realize automatic solution and calculation of each process, with fewer manual operation steps, which is conducive to reducing labor intensity, and the method can be applied to different operation and maintenance conditions, with good scalability. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0050] Figure 1A flow chart of a method provided by an embodiment of the present invention;

[0051] Figure 2 This is a schematic diagram of the amplitude measurement process;

[0052] Figure 3 It is a schematic diagram of the phase measurement process;

[0053] Figure 4 This is a circuit block diagram of an amplitude / phase measuring instrument;

[0054] Figure 5 This is a schematic diagram of the model training process. DETAILED DESCRIPTION

[0055] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0056] The purpose of the present invention is to provide a method for detecting the thickness of the PRTV coating on an insulator, which does not require complex true-type tests, saves test and labor costs, is simple and convenient, can be applied to different operation and maintenance conditions, and has good scalability.

[0057] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0058] The technical solution adopted by the present invention is to utilize the fact that when electromagnetic waves propagate to the material under test, the amplitude and phase will change, and polarization will occur. Due to the different thicknesses of the coatings on the tested pieces, the above parameters will change to different degrees. The functional relationship between the parameter change and the coating thickness is used to measure the PRTV coating thickness, and the relationship curve between the parameter change and the coating thickness is processed with the help of deep learning.

[0059] Figure 1 A flow chart of a method provided by an embodiment of the present invention, such as Figure 1 As shown, the present invention provides a method for detecting the thickness of an insulator PRTV coating, comprising:

[0060] Step 100: measuring the amplitude and phase data of the PRTV coating of the insulator by using an amplitude-phase method;

[0061] The amplitude phase method of coating thickness measurement is based on physical principles, especially those related to vibration, fluctuation and electromagnetic field. When the measuring probe contacts the coating to be measured, specific vibration or electromagnetic field will be generated. These vibrations or electromagnetic fields are directly related to the thickness of the coating. By measuring and analyzing the amplitude and phase changes of these vibrations or electromagnetic fields, the thickness of the coating can be inferred.

[0062] Step 200: pre-processing the measured amplitude and phase data, and establishing an image curve of the relationship between the thickness of the PRTV coating at different positions and the amplitude and phase parameters according to the corresponding insulator thickness;

[0063] Step 300: constructing an insulator PRTV coating thickness detection model, and using an image curve of the relationship between the thickness at different positions of the PRTV coating and the amplitude phase parameters as a training set, and training the model based on the training set;

[0064] Step 400: Evaluate model performance using validation data independent of the training set and adjust hyperparameters;

[0065] Step 500: Optimizing the model;

[0066] Step 600: deploy the optimized model and perform thickness detection on the PRTV coating of the insulator to be tested.

[0067] In step 100, the amplitude and phase data of the insulator PRTV coating are measured by the amplitude phase method, specifically:

[0068] Step 101: perform amplitude measurement to obtain amplitude data;

[0069] Step 102: Perform phase measurement to obtain phase data.

[0070] like Figure 2 As shown, in step 101, amplitude measurement is performed to obtain amplitude data, specifically:

[0071] According to the characteristics of the coating to be measured and the measurement requirements, select an amplitude measuring instrument with appropriate frequency range and measurement accuracy;

[0072] Calibrate the selected amplitude measuring instrument to ensure its accuracy and reliability. The calibration process usually includes zero point calibration and sensitivity calibration to ensure that the output value of the instrument at different amplitudes is proportional to the actual amplitude;

[0073] According to the material and thickness of the coating to be measured, select a suitable amplitude measurement probe, install it on the amplitude measurement instrument, and ensure that it is stable and firm;

[0074] Clean the sample to be tested, remove impurities such as oil, dust, etc., and ensure that the surface of the sample to be tested is flat and smooth without scratches or dents;

[0075] According to the material and thickness of the coating to be measured, set the parameters of the amplitude measuring instrument, such as measurement frequency, measurement range, etc., to ensure that the measurement parameter settings meet the measurement requirements;

[0076] Place the amplitude measurement probe on the surface of the sample to be measured, ensure that the probe is in close contact with the sample to be measured, and adjust the position and direction of the probe to obtain the best measurement effect;

[0077] Start the amplitude measuring instrument, start measuring the amplitude information of the measured coating, wait for the measuring instrument to complete the measurement, and record the measurement results to ensure that the recorded data is accurate and complete, and obtain the amplitude data of the measured coating;

[0078] Filter and smooth the measured amplitude data to remove noise and interference and improve the accuracy and reliability of the data;

[0079] An amplitude spectrum is constructed based on the processed amplitude data, and the amplitude spectrum can reflect the amplitude response of the measured covering at different frequencies;

[0080] According to the changing trend and characteristics of the amplitude spectrum, combined with the material and characteristics of the measured coating, the thickness of the measured coating can be preliminarily inferred.

[0081] like Figure 3 As shown, in step 102, phase measurement is performed to obtain phase data, specifically:

[0082] According to the characteristics of the coating to be measured and the measurement requirements, select a phase measurement instrument with appropriate frequency range and measurement accuracy;

[0083] Calibrate selected phase measurement instruments;

[0084] According to the material and thickness of the coating to be measured, select a suitable phase measurement probe and install it on the phase measurement instrument;

[0085] Clean the sample to be tested;

[0086] According to the material and thickness of the coating to be measured, set the parameters of the phase measurement instrument, such as measurement frequency, measurement range, etc. Ensure that the measurement parameter settings meet the measurement requirements in order to accurately capture the phase information;

[0087] Place the phase measurement probe on the surface of the sample to be measured, ensure that the probe is in close contact with the sample to be measured, and adjust the position and direction of the probe to obtain the best measurement effect;

[0088] Start the phase measuring instrument, start measuring the phase information of the measured coating, wait for the measuring instrument to complete the measurement, and record the measurement results to ensure that the recorded data is accurate and complete, including key information such as phase difference and frequency, and obtain the phase data of the measured coating;

[0089] Filtering and smoothing the measured phase data;

[0090] A phase spectrum is constructed based on the processed phase data, and the phase spectrum can reflect the phase response of the measured coating at different frequencies;

[0091] By comparing the measured phase data with the reference phase information (such as the phase information of the coating with known thickness), the phase difference is calculated. The phase difference is directly related to the thickness of the coating being measured and can be converted through a mathematical model or algorithm;

[0092] According to the phase difference and the known mathematical model or algorithm, based on the phase difference and in combination with the material and characteristics of the measured coating, the thickness of the measured coating is preliminarily inferred.

[0093] Among them, the circuit block diagram of the amplitude / phase measurement instrument is as follows Figure 4 As shown;

[0094] Combining the amplitude spectrum with the phase spectrum can more accurately determine the thickness of the coating. This method utilizes the complementarity between amplitude and phase to improve the accuracy and reliability of the measurement.

[0095] In step 200, the measured amplitude and phase data are preprocessed, and an image curve of the relationship between the thickness of the PRTV coating at different positions and the amplitude and phase parameters is established according to the corresponding insulator thickness, specifically:

[0096] The measured amplitude data and the thickness of the measured coating inferred from the amplitude data, the measured phase data and the thickness of the measured coating inferred from the phase data are obtained, and data cleaning (noise, outliers and missing values ​​are removed) and standardization or normalization are performed on them. Based on the processed data, an image curve of the relationship between the thickness of the PRTV coating at different positions and the amplitude and phase parameters is generated.

[0097] In step 300, an insulator PRTV coating thickness detection model is constructed, and an image curve based on the relationship between the thickness at different positions of the PRTV coating and the amplitude phase parameter is used as a training set. The model is trained based on the training set, specifically:

[0098] Define layers and nodes, stack various types of layers to form a deep network, build an insulator PRTV coating thickness detection model based on the deep network, and use the image curve based on the relationship between the thickness and amplitude phase parameters of the PRTV coating at different positions as the training set, and train the model based on the training set. The specific training process is as follows: Figure 5 As shown, introduce it:

[0099] The model training process first sets random weights and biases, initializes the parameters, inputs data, calculates the predicted results through the network, compares the predicted results with the true labels, obtains the loss value, and then updates the network parameters according to the gradient descent method to minimize the loss function. The above process is repeated many times (epochs) until convergence. The process can be summarized as parameter initialization, forward propagation, loss calculation, back propagation, and training iterations.

[0100] Step 400: evaluate model performance through validation data independent of the training set and adjust hyperparameters to prevent overfitting;

[0101] In step 500, the model is optimized by adjusting hyperparameters such as learning rate, batch size, and optimizer to improve model performance;

[0102] In step 600, the optimized model is deployed (the trained model is saved as a file and then loaded into the production environment or API service, and new data is predicted or processed in real time), and the thickness of the PRTV coating of the insulator to be tested is detected.

[0103] The core of this invention is to combine microwave detection technology with deep learning algorithms, using the reflection, attenuation and scattering characteristics of microwave propagation, combined with the convolutional neural network (CNN) in deep learning to extract and analyze the features of electromagnetic wave signals, and accurately detect the thickness of PRTV coatings. This combination not only improves the detection accuracy, but also overcomes the limitations of traditional detection methods (such as infrared, ultrasonic, etc.) on complex surfaces or irregular objects.

[0104] The present invention uses deep learning algorithms to automatically extract complex patterns from test data, optimize the data analysis process, and reduce human errors. Compared with traditional coating measurements, the present invention can achieve automated analysis through deep learning models, improve detection efficiency and accuracy, and meet large-scale detection needs.

[0105] The present invention also takes into account the errors caused by environmental factors affecting detection, and corrects the changes in microwave signals through deep learning model training and adaptive algorithms, thereby ensuring the consistency and reliability of detection results in different environments.

[0106] The present invention has the following advantages:

[0107] (1) Low energy consumption and easy operation:

[0108] The microwave method has the characteristics of low energy consumption and fast response, can achieve instantaneous measurement, and does not require the use of coupling agents, which reduces complex preparation work and improves detection efficiency. At the same time, microwave technology can measure coatings on complex and irregular surfaces, overcoming the limitations of traditional technologies such as ultrasound and infrared on such surfaces.

[0109] (2) Non-contact detection reduces the risk of injury:

[0110] Compared with traditional detection methods that require physical contact, the present invention uses microwave technology for non-contact detection, avoiding scratches, indentations or other forms of physical damage to the PRTV coating. This not only helps to protect the coating integrity of power equipment, but also improves the safety and reliability of detection.

[0111] (3) Deep learning improves detection accuracy:

[0112] The present invention combines deep learning algorithms, especially convolutional neural networks (CNN), to automatically extract coating thickness features from complex electromagnetic wave data. Through automated processing of changes in microwave signal reflection, attenuation and other parameters, the present invention achieves high-precision identification of coating thickness and significantly reduces detection errors.

[0113] (4) Intelligent and automated testing process:

[0114] Deep learning technology makes the detection process highly automated, reducing the need for human intervention. It not only improves detection efficiency, but also maintains consistency during large-scale detection, greatly reduces the uncertainty caused by human errors and subjective judgments, and ensures the objectivity and repeatability of the detection results.

[0115] (5) Strong adaptability and broad application prospects:

[0116] The present invention has good adaptability to complex environments and can effectively cope with the influence of external factors such as different temperatures and humidity, ensuring the stability and consistency of the detection results. At the same time, the adaptive characteristics of deep learning enable it to continuously optimize its own detection capabilities, thereby adapting to different types of PRTV coatings and various coating thickness requirements, and has wide application potential under complex field conditions.

[0117] (6) Support big data analysis and fault prediction:

[0118] Due to the application of deep learning algorithms, the present invention can extract valuable information from a large amount of test data, which is not only used for real-time detection of coating thickness, but also can provide strong technical support for coating quality control, production optimization and future fault prediction.

[0119] The microwave detection technology of the present invention can also be replaced by technologies of other electromagnetic wave frequency bands, such as radio frequency detection technology or millimeter wave technology, which can also measure the thickness of the coating through the interaction between its wavelength and the coating thickness. These technologies are similar to the microwave method in principle, and the thickness is inferred through the changing characteristics of electromagnetic waves such as reflection, attenuation, and scattering, but the operating frequencies are different, and a more suitable frequency band can be selected according to specific needs.

[0120] In addition to using convolutional neural networks (CNN) in deep learning for data processing and feature extraction, other machine learning algorithms, such as support vector machines (SVM), random forests or decision trees, can also process microwave signal reflection data and establish the relationship between coating thickness and electromagnetic wave characteristic parameters. These alternative data processing methods can also achieve accurate detection of PRTV coating thickness, especially in specific data sets, which may have higher efficiency or better robustness.

[0121] In addition to using traditional signal analysis methods or deep learning algorithms to process microwave signals, the present invention can also decompose and analyze the reflected microwave signals through signal processing techniques such as spectrum analysis, wavelet transform or Fourier transform. These alternative signal processing methods may be more effective in different frequency bands or complex signal environments, and can help inspectors more intuitively judge the thickness changes of PRTV coatings.

[0122] The present invention proposes to perform coating thickness detection under normal temperature and pressure, but in actual operation, it is also possible to consider performing detection under different environmental conditions to improve accuracy. For example, other air pressure environments or low temperature environments can reduce the impact of environmental interference on microwave signal propagation, thereby further improving the accuracy and stability of thickness detection.

[0123] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0124] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A method for detecting the thickness of an insulator PRTV coating, characterized in that: include: The amplitude and phase data of the insulator PRTV coating are measured by the amplitude phase method; The measured amplitude and phase data are preprocessed, and the image curve of the relationship between the thickness of the PRTV coating at different positions and the amplitude and phase parameters is established according to the corresponding insulator thickness; The insulator PRTV coating thickness detection model is constructed, and the image curves based on the relationship between the thickness of the PRTV coating at different positions and the amplitude phase parameters are used as the training set, and the model is trained based on the training set; Evaluate model performance and adjust hyperparameters using validation data independent of the training set; Optimize the model; The optimized model is deployed to detect the thickness of the PRTV coating on the insulator to be tested.

2. The method according to claim 1, characterized in that: The amplitude and phase data of the insulator PRTV coating are measured by the amplitude phase method, specifically: Perform amplitude measurement and obtain amplitude data; Perform phase measurement and obtain phase data.

3. The method according to claim 2, characterized in that Perform amplitude measurement and obtain amplitude data, specifically: According to the characteristics of the coating to be measured and the measurement requirements, select an amplitude measuring instrument with appropriate frequency range and measurement accuracy; Calibrate the selected amplitude measuring instrument; According to the material and thickness of the coating to be measured, select a suitable amplitude measurement probe and install it on the amplitude measurement instrument; Clean the sample to be tested; Set the parameters of the amplitude measuring instrument according to the material and thickness of the coating to be measured; Place the amplitude measurement probe on the surface of the sample to be measured, ensure that the probe is in close contact with the sample to be measured, and adjust the position and direction of the probe to obtain the best measurement effect; Start the amplitude measuring instrument to measure the amplitude data of the coating under test; Filtering and smoothing the measured amplitude data; constructing an amplitude spectrum based on the processed amplitude data; According to the changing trend and characteristics of the amplitude spectrum, combined with the material and characteristics of the measured coating, the thickness of the measured coating can be preliminarily inferred.

4. The method according to claim 2, characterized in that: Perform phase measurement and obtain phase data, specifically: According to the characteristics of the coating to be measured and the measurement requirements, select a phase measurement instrument with appropriate frequency range and measurement accuracy; Calibrate selected phase measurement instruments; According to the material and thickness of the coating to be measured, select a suitable phase measurement probe and install it on the phase measurement instrument; Clean the sample to be tested; Set the parameters of the phase measuring instrument according to the material and thickness of the coating to be measured; Place the phase measurement probe on the surface of the sample to be measured, ensure that the probe is in close contact with the sample to be measured, and adjust the position and direction of the probe to obtain the best measurement effect; Start the phase measuring instrument to measure the phase data of the measured coating; Filtering and smoothing the measured phase data; constructing a phase spectrum based on the processed phase data; By comparing the measured phase data with the reference phase information, the phase difference is calculated; Based on the phase difference, combined with the material and characteristics of the measured coating, the thickness of the measured coating can be preliminarily inferred.

5. The method according to claim 1, characterized in that The measured amplitude and phase data are preprocessed, and the image curve of the relationship between the thickness of the PRTV coating at different positions and the amplitude and phase parameters is established according to the corresponding insulator thickness. Specifically: The measured amplitude data and the thickness of the measured coating inferred from the amplitude data, the measured phase data and the thickness of the measured coating inferred from the phase data are obtained, and data cleaning and standardization or normalization processing are performed on them. Based on the processed data, an image curve of the relationship between the thickness of the PRTV coating at different positions and the amplitude and phase parameters is generated.

6. The method according to claim 1, characterized in that The insulator PRTV coating thickness detection model is constructed, and the image curve based on the relationship between the thickness of the PRTV coating at different positions and the amplitude phase parameters is used as the training set. The model is trained based on the training set, specifically: Define layers and nodes, stack various types of layers to form a deep network, build an insulator PRTV coating thickness detection model based on the deep network, and use the image curve of the relationship between the thickness and amplitude phase parameters of the PRTV coating at different positions as the training set, and train the model based on the training set.

Images