Hygrothermal aging detection method of unsaturated polyester resin material and related device
Through the method based on color feature extraction and machine learning, the existing UPR insulating material aging detection methods are solved, and the rapid and accurate detection of the aging degree of UPR materials is achieved, and the power system needs to monitor and predict the insulation status in real time.
Patent Information
- Application Number
- CN202510069866.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-01-16
- Publication Date
- 2025-05-13
AI Technical Summary
The existing aging detection methods for unsaturated polyester resin (UPR) insulating materials have problems such as low detection efficiency, insufficient accuracy, strong subjectivity and high equipment requirements, which are difficult to meet the needs of modern power systems for real-time monitoring and predictive insulation status.
Using detection methods based on color feature extraction and machine learning, the color feature vectors of the material are obtained, and the aging degree prediction is predicted using support vector machine regression, random forest regression or multi-layer perceptron model to achieve rapid, accurate and objective detection of the aging degree of UPR materials.
It realizes rapid, accurate and objective detection of the aging degree of UPR materials, can meet the online monitoring and predictive requirements of power equipment insulation status, reduces detection costs, improves cost-effectiveness, and can promptly detect and prevent the aging trend of insulating materials, and ensure the safe operation of the power system.
Smart Images
Figure CN119985950A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of power system material aging status monitoring, and in particular relates to a wet heat aging detection method for unsaturated polyester resin materials and a related device. Background Art
[0002] Unsaturated Polyester Resin (UPR) is an important thermosetting polymer material with excellent mechanical properties, electrical insulation properties, chemical corrosion resistance, heat and fire resistance, and easy processing and molding. It is widely used in the fields of power equipment insulation components, electronic packaging materials, composite materials, and building materials. Explanatoryally, in power systems, UPR is often used to manufacture key components such as insulators, transformer winding insulation, generator slot wedges, cable trough boxes, etc.; during long-term operation, UPR insulation materials will be affected by electrical stress, thermal stress, mechanical stress, and environmental factors (for example, humidity, ultraviolet rays, air pollution, etc.). Further explanatory, the damp and heat aging in actual application fields change The phenomenon is one of the main factors leading to the aging and deterioration of UPR insulation materials; damp heat aging refers to the deterioration process of the physical, chemical and mechanical properties of the material caused by the combined action of moisture and heat energy when the material is exposed to high temperature and high humidity environment for a long time. This process usually involves the breakage of molecular chains, changes in cross-linking degree, hygroscopic expansion and reorganization of the microstructure inside the material, resulting in reduced strength, reduced toughness, and poor dimensional stability of the material. These physical and chemical changes will cause the dielectric properties, fire resistance and mechanical properties of the material to decline, ultimately affecting the safe operation of power equipment.
[0003] At present, the existing traditional UPR insulation material aging detection methods mainly include dielectric loss factor measurement, volume resistivity measurement, glass transition temperature (Tg) test, infrared spectroscopy analysis (FTIR), scanning electron microscope (SEM) observation, etc. Although the above existing methods can reflect the aging degree of the material to a certain extent, they still have the following defects, including: (1) Low detection efficiency: It takes a lot of time and manpower to prepare, test and analyze samples, and it is impossible to achieve real-time monitoring of insulation materials of power equipment; (2) Limited accuracy: Some test methods are not sensitive enough to small changes in the early aging stage, making it difficult to detect potential insulation failures in a timely manner; (3) High subjectivity: The analysis and interpretation of test results depends on the experience and professional level of the testers, and there is a certain degree of subjective bias; (4) High equipment requirements: expensive professional testing equipment is required, which increases the testing cost and is not conducive to widespread application; In summary, an efficient, accurate and objective UPR insulation material aging detection method is urgently needed to meet the requirements of modern power systems for real-time monitoring and predictiveness of insulation status. Summary of the invention
[0004] The purpose of the present invention is to provide a method and related device for detecting the hygrothermal aging of unsaturated polyester resin materials to solve one or more of the above-mentioned technical problems. The technical solution provided by the present invention is a method for detecting the hygrothermal aging of unsaturated polyester resin materials based on color feature extraction and machine learning, which can realize rapid, accurate and objective detection of the aging degree of UPR materials and meet the needs of online monitoring and predictability of material status.
[0005] In order to achieve the above object, the present invention adopts the following technical solutions: In a first aspect, the present invention provides a method for detecting wet heat aging of an unsaturated polyester resin material, comprising the following steps: Obtaining a color feature vector of the unsaturated polyester resin material to be detected; Based on the acquired color feature vector, the trained aging detection model is used to make predictions to obtain the aging degree prediction value; Compare the obtained aging degree prediction value with the preset aging degree threshold to obtain the aging degree detection and grading result; in, The aging detection model adopts a support vector machine regression model, a random forest regression model or a multi-layer perceptron model; The training steps of the aging detection model include: Acquire a training sample set; wherein the training sample set includes color feature vectors of unsaturated polyester resin material samples of different aging degrees and aging degree labels corresponding to each color feature vector; the aging degree label corresponding to each color feature vector uses the relative change rate of the dielectric loss factor as an aging degree indicator; The aging detection model is trained using the training sample set, and the model parameters are adjusted to obtain a trained aging detection model after reaching a preset convergence condition.
[0006] A further improvement of the present invention is that the step of obtaining the color feature vector of the unsaturated polyester resin material to be detected comprises: Acquire an RGB image of the unsaturated polyester resin material to be detected and convert it into a Lab color space to obtain a converted image; Based on the converted image, the global statistical features and color histogram features of each color channel are obtained to form a color feature vector of the unsaturated polyester resin material to be detected.
[0007] A further improvement of the present invention is that the step of obtaining a training sample set includes: Prepare multiple unsaturated polyester resin material samples with consistent size, shape and quality; Placing the prepared multiple unsaturated polyester resin material samples in an aging test box, applying different temperatures, humidity and aging time to perform wet heat aging treatment, and obtaining multiple unsaturated polyester resin material samples with different aging degrees; Acquire RGB images of unsaturated polyester resin material samples with different aging degrees and convert them into Lab color space to obtain multiple converted images; based on each converted image, obtain global statistical features and color histogram features of each color channel to form color feature vectors of unsaturated polyester resin material samples with different aging degrees; Based on multiple unsaturated polyester resin material samples with different aging degrees, an aging degree label corresponding to each color feature vector is obtained; wherein the aging degree label uses the relative change rate of the dielectric loss factor as the aging degree indicator, and the relative change rate The calculation expression is: ; In the formula, is the dielectric loss factor of the sample after aging treatment; is the dielectric loss factor of the sample without aging treatment.
[0008] A further improvement of the present invention is that the global statistical features include mean value, standard deviation, skewness and kurtosis; and the color histogram features include entropy, energy and contrast of the histogram.
[0009] A further improvement of the present invention is that, in the step of obtaining the aging degree label corresponding to each color feature vector based on a plurality of unsaturated polyester resin material samples with different aging degrees, the dielectric loss factor of the sample without aging treatment and the sample after aging treatment are measured and obtained at room temperature and 50 Hz.
[0010] A further improvement of the present invention lies in that, in the step of placing the prepared multiple unsaturated polyester resin material samples in an aging test box, applying different temperatures, humidity and aging times to perform wet heat aging treatment, and obtaining multiple unsaturated polyester resin material samples with different aging degrees, the humidity ranges from 50% to 95%, the temperature ranges from 50°C to 90°C, and the aging time ranges from 168h to 840h.
[0011] A further improvement of the present invention is that the step of using the training sample set to train the aging detection model, adjusting the model parameters, and obtaining the trained aging detection model after reaching a preset convergence condition comprises: Dividing the training sample set into a training set and a validation set; Based on the training set, the aging detection model is trained, model parameters are adjusted, and a trained aging detection model is obtained; Based on the validation set, a prediction performance evaluation is performed on the trained aging detection model, and when the evaluation result meets the preset performance requirement, a trained aging detection model is obtained; Among them, the indicators used in the prediction performance evaluation include mean absolute error, mean square error and determination coefficient.
[0012] In a second aspect, the present invention provides a system for detecting wet heat aging of unsaturated polyester resin materials, comprising: A color feature vector acquisition module, used to acquire the color feature vector of the unsaturated polyester resin material to be detected; A prediction module is used to make predictions based on the acquired color feature vectors using a trained aging detection model to obtain an aging degree prediction value; A comparison module, used to compare the obtained aging degree prediction value with a preset aging degree threshold value to obtain an aging degree detection and classification result; in, The aging detection model adopts a support vector machine regression model, a random forest regression model or a multi-layer perceptron model; The training steps of the aging detection model include: Acquire a training sample set; wherein the training sample set includes color feature vectors of unsaturated polyester resin material samples of different aging degrees and aging degree labels corresponding to each color feature vector; the aging degree label corresponding to each color feature vector uses the relative change rate of the dielectric loss factor as an aging degree indicator; The aging detection model is trained using the training sample set, and the model parameters are adjusted to obtain a trained aging detection model after reaching a preset convergence condition.
[0013] In a third aspect of the present invention, there is provided an electronic device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method for detecting wet heat aging of unsaturated polyester resin materials as described in any one of the first aspect of the present invention is implemented.
[0014] In a fourth aspect, the present invention provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for detecting wet heat aging of unsaturated polyester resin materials as described in any one of the first aspects of the present invention.
[0015] Compared with the prior art, the present invention has the following beneficial effects: In order to solve the problems of low detection efficiency, insufficient accuracy and strong subjectivity existing in the existing technical solutions, the present invention specifically discloses a method for detecting wet heat aging of unsaturated polyester resin (UPR) materials based on color feature extraction and machine learning. The method utilizes the changes in color characteristics of the material surface, combines with the quantitative evaluation of the degree of aging, and adopts a machine learning algorithm to establish a mapping model between color characteristics and the degree of aging. The method can realize fast, accurate and objective detection of the degree of aging of UPR materials, and can meet the needs of online monitoring and predictiveness of the insulation status of power equipment.
[0016] Specifically, the method for detecting the wet heat aging of unsaturated polyester resin materials provided by the present invention has the following significant advantages, including: (1) High efficiency: Through digital image processing and machine learning algorithms, the aging degree of UPR materials can be quickly detected, reducing the cumbersome testing process of traditional methods; (2) Accuracy: The accuracy of aging prediction is improved by using multi-dimensional color features to capture subtle changes on the material surface and combining them with the powerful nonlinear fitting capability of the machine learning model. (3) Objectivity: It avoids the subjectivity and instability of manual testing, and the test results are highly repeatable and reliable; (4) Economical: No expensive professional testing equipment is required, which reduces the testing cost and has a high cost-effectiveness; (5) Safety: It can timely detect and predict the aging trend of insulation materials, prevent the occurrence of power equipment failures, and ensure the safe operation of the power system. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below; obviously, the drawings described below are some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0018] Figure 1 It is a schematic flow chart of a method for detecting wet heat aging of an unsaturated polyester resin material in an embodiment of the present invention; Figure 2 It is a flow chart of a method for detecting wet heat aging of unsaturated polyester resin materials based on color feature extraction and machine learning in a specific embodiment of the present invention; Figure 3 It is a schematic diagram of a wet heat aging detection system for unsaturated polyester resin materials in an embodiment of the present invention. DETAILED DESCRIPTION
[0019] In order to make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions in the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention; it is obvious that the described embodiments and technical solutions are only part of the embodiments of the present invention, not all of the embodiments.
[0020] All other embodiments obtained by those of ordinary skill in the art without creative work based on the technical solutions disclosed in the embodiments of the present invention belong to the scope of protection of the present invention. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device including a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0021] See also Figure 1 The embodiment of the present invention discloses a method for detecting wet heat aging of an unsaturated polyester resin material, comprising the following steps: Step 1, obtaining a color feature vector of the unsaturated polyester resin material to be detected; Step 2, based on the color feature vector obtained in step 1, use the trained aging detection model to make a prediction to obtain an aging degree prediction value; Step 3, comparing the aging degree prediction value obtained in step 2 with a preset aging degree threshold to obtain an aging degree detection and grading result; in, The aging detection model adopts a support vector machine regression model, a random forest regression model or a multi-layer perceptron model; The training steps of the aging detection model include: Acquire a training sample set; wherein the training sample set includes color feature vectors of unsaturated polyester resin material samples of different aging degrees and their corresponding aging degree labels; when acquiring the aging degree labels, the relative change rate of the dielectric loss factor is used as the aging degree indicator; The aging detection model is trained using the training sample set, and the model parameters are adjusted to obtain a trained aging detection model after reaching a preset convergence condition.
[0022] The technical solution disclosed in the embodiment of the present invention cleverly combines digital image processing and machine learning algorithms to achieve rapid detection of the aging degree of UPR materials; compared with the cumbersome and time-consuming test process of the traditional method, the technical solution of the present invention significantly improves the detection efficiency, can complete the analysis of a large number of samples in a short time, and greatly saves time and labor costs. The technical solution of the present invention adopts multidimensional color feature capture technology, which can accurately capture subtle changes on the surface of the material, which are often difficult to detect with the naked eye; at the same time, combined with the powerful nonlinear fitting ability of the machine learning model, the degree of aging is accurately predicted, which greatly improves the accuracy of the detection. This high-precision detection method provides a scientific basis for the maintenance and replacement of materials. The technical solution of the present invention is completely based on data and algorithms for detection, avoiding the subjectivity and instability of manual detection. The detection results are highly repeatable and reliable. No matter who performs the operation, consistent results can be obtained. This objective detection method provides a fair and accurate basis for the evaluation of materials. The technical solution of the present invention does not require expensive professional testing equipment, but only requires ordinary image acquisition equipment and a computer to complete the detection, which greatly reduces the detection cost and improves the cost performance. For occasions where a large number of test samples are required, the economy of the technical solution is particularly prominent. The technical solution of the present invention can timely discover and predict the aging trend of insulating materials, providing strong support for the maintenance of power equipment. Through regular detection, the occurrence of power equipment failures can be prevented and the safe operation of the power system can be guaranteed. This is of great significance for improving the reliability of power equipment, extending its service life and ensuring the stable operation of the power grid.
[0023] See also Figure 2 The embodiment of the present invention discloses a method for detecting the hygrothermal aging of an unsaturated polyester resin material, specifically a method for detecting the hygrothermal aging of an unsaturated polyester resin material based on color feature extraction and machine learning, comprising the following steps: Step 1, sample preparation and wet heat aging treatment, including: Step 1.1, sample preparation, including: preparing unsaturated polyester resin material samples with consistent size and shape, ensuring the consistency of sample quality (such as smooth surface, no bubbles, no impurities, etc.); In a specific exemplary technical solution, commercial unsaturated polyester resin can be selected, and an appropriate amount of initiator (such as methyl ethyl ketone peroxide) and accelerator (such as cobalt salt) can be added, and the mixture is thoroughly stirred at room temperature; then, the mixed material is poured into a mold, cured at room temperature for 24 hours, and then post-cured at 80°C for 2 hours to obtain a dense unsaturated polyester resin material sample.
[0024] Step 1.2, subjecting the sample to different temperatures, humidity and aging time in an aging test chamber to perform wet heat aging treatment; In a specific exemplary technical solution, the wet heat aging treatment includes: The prepared multiple unsaturated polyester resin material samples were placed in an aging test chamber; different temperatures (50°C, 60°C, 70°C, 80°C, 90°C), humidity (50%, 65%, 80%, 95%) and aging time (168 hours, 280 hours, 420 hours, 560 hours, 840 hours) were applied for wet heat aging treatment to obtain multiple unsaturated polyester resin material samples with different aging degrees. During the aging process, the temperature and humidity parameters in the test chamber were strictly controlled to ensure that each sample was uniformly aged under the specified conditions. For example, the humidity range of the wet heat aging treatment was 50%~95%, the temperature range was 50℃~90℃, and the aging time range was 168 hours to 840 hours.
[0025] Step 2, aging assessment, includes: Step 2.1, electrical performance test, including: using a dielectric loss meter (such as a Schering bridge) to measure the dielectric loss factor of the sample at room temperature and 50 Hz and dielectric constant ; Before testing, the sample can be dried in a drying oven at 50°C for 24 hours to eliminate the influence of moisture.
[0026] Step 2.2, aging calculation, including: relative change rate of dielectric loss factor As an indicator of aging, the calculation formula is: ; In the formula, is the dielectric loss factor of the sample after aging; is the dielectric loss factor of the fresh sample without aging treatment.
[0027] Step 3, image acquisition and preprocessing, includes: Step 3.1, RGB (Red, Green, Blue) image acquisition, including: photographing the surface of the aged sample using a high-resolution digital camera (e.g., with a pixel of not less than 12 million) under constant light source conditions (D65 standard light source with a color temperature of 5500K and an illumination of 1000±50 lux); In the specific exemplary technical solution, the camera parameters are fixed, the ISO sensitivity is 100, the aperture value is F / 8, and the shutter speed is adjusted according to the exposure compensation to ensure that the image exposure is moderate. The shooting distance and angle are fixed, and the lens is perpendicular to the sample surface at a distance of 50 cm.
[0028] Step 3.2, image preprocessing, includes: using image processing software (for example, Matlab, Python's OpenCV library, etc.) to crop, rotate and normalize the acquired images to unify the image size (such as 512×512 pixels); performing color correction, using a standard color card to adjust the white balance, and reducing the influence of uneven lighting and ambient light.
[0029] Step 4, color feature extraction, includes: Step 4.1, color space conversion, includes: converting the preprocessed RGB image into Lab color space. The L (brightness), a (red-green axis), and b (yellow-blue axis) components of the Lab color space can better reflect the color differences perceived by the human eye.
[0030] Step 4.2, characteristic parameter calculation, includes: For each color channel (L, a, b), calculate the global statistical features, and the calculation expressions are: Mean: ; Standard Deviation: ; Skewness: ; Kurtosis: ; in, is the color value of the i-th pixel; is the total number of pixels.
[0031] Step 4.3, extracting color histogram features, including: dividing the color value of each channel into 256 gray levels, counting the number of pixels at each gray level to obtain a histogram, calculating the entropy, energy, contrast and other texture features of the histogram, and finally forming a color feature vector containing multiple feature dimensions.
[0032] Step 5, constructing a training sample set, including: taking the color feature vectors of samples with different aging degrees as input variables X, and the corresponding aging degrees As the output label Y, construct the training sample set and the test sample set; In order to avoid overfitting, the cross-validation method is used to divide the sample set into a training set and a validation set (such as a 7:3 or 8:2 ratio).
[0033] Step 6, aging detection model training and verification, including: Step 6.1, model selection, includes: selecting an appropriate machine learning algorithm, such as Support Vector Regression (SVR), Random Forest Regression (RFR), Multilayer Perceptron (MLP), etc.
[0034] Step 6.2, model training, includes: using the training sample set to train the selected model and adjust the model parameters; for SVR, select a kernel function (such as radial basis function RBF), adjust the penalty parameter C and kernel function parameters ; For RFR, adjust parameters such as the number of decision trees (n_estimators), maximum depth (max_depth), and minimum number of sample splits (min_samples_split); for MLP, design the network structure (number of hidden layers and neurons), select the activation function and optimization algorithm.
[0035] Step 6.3, model validation, includes: using the validation set to evaluate the predictive performance of the model and calculating the following metrics: Mean Absolute Error (MAE): ; Mean Squared Error (MSE): ; Coefficient of Determination (R 2 ): ; in, is the true value, is the predicted value, is the average value of the true value, and n is the number of samples; The model and parameters with the best performance can be selected based on the verification results.
[0036] Step 7, aging degree detection and evaluation, including: for the UPR to be tested, collect images according to the above method, extract color features, and input the color feature vector into the trained aging detection model to predict the aging degree prediction value , the aging degree is classified according to the aging degree threshold; In an exemplary technical solution, slight aging: ; Moderate aging: ; Severe aging: ; Furthermore, corresponding maintenance and repair suggestions can be put forward according to the degree of aging.
[0037] With the rapid development of artificial intelligence and digital image processing technology, image-based material surface feature extraction and analysis methods have shown broad application prospects in the field of material aging detection. During the aging process of the material, the macroscopic and microscopic structures of its surface will change, especially the changes in color and texture features are more obvious. These changes can be acquired through high-resolution image acquisition equipment, combined with machine learning algorithms, to establish a quantitative relationship between the surface features of the material and the degree of aging, thereby achieving rapid and accurate detection of the degree of material aging. Based on the above principles, an embodiment of the present invention specifically discloses a method for detecting the degree of aging of an unsaturated polyester resin material under the action of a humid and hot environment based on a combination of color feature extraction and a machine learning algorithm. Figure 2 The whole process from sample preparation, wet heat aging treatment, aging assessment, image acquisition and preprocessing, color feature extraction, model training and verification to aging degree detection and evaluation is demonstrated.
[0038] In a specific embodiment of the present invention, Step 1: Sample preparation and heat aging treatment process includes: Sample preparation: Commercial unsaturated polyester resin (model UPR-196) was selected, 1% (mass fraction) of methyl ethyl ketone peroxide (MEKP) was added as an initiator, and 0.5% (mass fraction) of cobalt cyclopentaneate was added as an accelerator, and the mixture was stirred thoroughly. The mixed resin was poured into a silicone mold with a size of 50 mm × 50 mm × 5 mm and cured at room temperature for 24 hours. The cured sample was post-cured in an oven at 80°C for 2 hours to ensure that the resin was completely cured.
[0039] Hygrothermal aging treatment: 10 groups of samples were prepared and subjected to different hygrothermal aging treatments; multiple unsaturated polyester resin material samples were placed in an aging test chamber and treated with different temperature, humidity and aging time conditions: for groups 1-3, the temperature was set to 50°C, the humidity was 50%, 65%, and 80%, and the aging time was 168 hours, 280 hours, and 420 hours, respectively; for groups 4-6, the temperature was set to 70°C, the humidity was 65%, 80%, and 95%, and the aging time was 168 hours, 280 hours, and 420 hours, respectively; for groups 7-9, the temperature was set to 90°C, the humidity was 50%, 80%, and 95%, and the aging time was 560 hours, 840 hours, and 560 hours, respectively; for group 10, the temperature was set to 80°C, the humidity was 95%, and the aging time was 840 hours. During the hygrothermal aging process, a temperature and humidity control system was used to monitor and record the temperature and humidity parameters in the aging test chamber in real time to ensure that each group of samples was uniformly aged under the specified conditions.
[0040] Step 2: The aging assessment process includes: Dielectric loss factor measurement: Use a Schering bridge (such as TH2829A) to measure the dielectric loss factor of the sample at room temperature and 50 Hz and dielectric constant . Each sample was measured 3 times and the average value was taken to ensure the reliability of the results; Aging degree calculation: Take the fresh sample (without aging treatment) as the reference and calculate the ; Result analysis: It was found that with the increase of humidity, temperature and aging time during wet heat aging treatment, There is a significant increase, indicating that the dielectric loss of the material has increased and the fire resistance has decreased.
[0041] Step 3: The process of image acquisition and preprocessing includes: RGB image acquisition: Use a Canon EOS 5D Mark IV camera with a standard 50 mm fixed-focus lens; shoot in a light box, the light source is an LED light with a color temperature of 5500 K, and the illumination is uniform. Camera settings: ISO 100, aperture F / 8, shutter speed 1 / 60 s, white balance set to manual mode; Image preprocessing: Use Adobe Photoshop software to crop and color correct the images; unify the image size to 512×512 pixels and save them in uncompressed TIFF format to avoid image quality loss.
[0042] Step 4: The process of color feature extraction includes: Color space conversion: In the Matlab environment, use the rgb2lab function to convert the RGB image to the Lab color space; Calculation of characteristic parameters: for each channel (L, a, b), the mean, standard deviation, skewness, and kurtosis were calculated, for a total of 12 characteristic parameters; the entropy, energy, contrast and other characteristics of the color histogram were extracted, for a total of 9 characteristic parameters; finally, the dimension of the color feature vector of each sample was 21.
[0043] Step 5: The process of constructing a training sample set includes: The color characteristics of 10 groups of aging samples (a total of 10 samples) and the corresponding A training sample set is formed; specifically, due to the small number of samples, the Leave-One-Out Cross Validation method is used for model training and verification.
[0044] Step 6: The process of aging detection model training and verification includes: Model selection: Select the support vector machine regression (SVR) model, and the kernel function is the radial basis function (RBF); Model training: Use the fitrsvm function of the support vector regression model for model training; Tuning Model Parameters: Penalty Coefficients , kernel function parameters ; Model validation: The leave-one-out validation method is used to calculate indicators such as MAE, MSE, and R2; The results show that the MAE is 1.5%, the MSE is 0.04, and the R2 is 0.98, indicating that the model has high prediction accuracy.
[0045] Step 7: The process of aging detection and evaluation includes: New sample detection: Image acquisition and color feature extraction of new samples that have not participated in model training.
[0046] Aging prediction: Input the color feature vector of the new sample into the trained SVR model to predict ; Result evaluation: Predicted results and actual measurements The difference is within 2%, which verifies the reliability of the model; According to the predicted , determine the aging degree of the sample and make maintenance recommendations.
[0047] The following are device embodiments of the present invention, which can be used to implement the method embodiments of the present invention. For details not disclosed in the device embodiments, please refer to the method embodiments of the present invention.
[0048] See also Figure 3 In an embodiment of the present invention, a wet heat aging detection system for unsaturated polyester resin materials is provided, comprising: A color feature vector acquisition module, used to acquire the color feature vector of the unsaturated polyester resin material to be detected; A prediction module is used to make predictions based on the acquired color feature vectors using a trained aging detection model to obtain an aging degree prediction value; A comparison module, used to compare the obtained aging degree prediction value with a preset aging degree threshold value to obtain an aging degree detection and classification result; in, The aging detection model adopts a support vector machine regression model, a random forest regression model or a multi-layer perceptron model; The training steps of the aging detection model include: Acquire a training sample set; wherein the training sample set includes color feature vectors of unsaturated polyester resin material samples of different aging degrees and aging degree labels corresponding to each color feature vector; the aging degree label corresponding to each color feature vector uses the relative change rate of the dielectric loss factor as an aging degree indicator; The aging detection model is trained using the training sample set, and the model parameters are adjusted to obtain a trained aging detection model after reaching a preset convergence condition.
[0049] In one embodiment of the present invention, a computer device is provided, the computer device includes a processor and a memory, the memory is used to store a computer program, the computer program includes program instructions, and the processor is used to execute the program instructions stored in the computer storage medium. The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, which is suitable for implementing one or more instructions, and is specifically suitable for loading and executing one or more instructions in the computer storage medium to implement the corresponding method flow or corresponding function; the processor described in the embodiment of the present invention can be used to perform the operation of the wet heat aging detection method of unsaturated polyester resin materials.
[0050] In one embodiment of the present invention, a storage medium is provided, specifically a computer-readable storage medium (Memory), which is a memory device in a computer device for storing programs and data. It is understandable that the computer-readable storage medium here can include both built-in storage media in the computer device and, of course, extended storage media supported by the computer device. The computer-readable storage medium provides a storage space, which stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by the processor are also stored in the storage space, and these instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM (Random Access Memory) memory, or a non-volatile memory (non-volatile memory), such as at least one disk memory. The processor can load and execute one or more instructions stored in the computer-readable storage medium to implement the corresponding steps of the wet heat aging detection method for unsaturated polyester resin materials in the above embodiment.
[0051] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, optical storage, etc.) containing computer-usable program codes.
[0052] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0053] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0054] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0055] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.
Claims
1. A method for detecting wet heat aging of unsaturated polyester resin materials, characterized in that: The following steps are involved: Obtaining a color feature vector of the unsaturated polyester resin material to be detected; Based on the acquired color feature vector, the trained aging detection model is used to make predictions to obtain the aging degree prediction value; Compare the obtained aging degree prediction value with the preset aging degree threshold to obtain the aging degree detection and grading result; in, The aging detection model adopts a support vector machine regression model, a random forest regression model or a multi-layer perceptron model; The training steps of the aging detection model include: Acquire a training sample set; wherein the training sample set includes color feature vectors of unsaturated polyester resin material samples of different aging degrees and aging degree labels corresponding to each color feature vector; the aging degree label corresponding to each color feature vector uses the relative change rate of the dielectric loss factor as an aging degree indicator; The aging detection model is trained using the training sample set, and the model parameters are adjusted to obtain a trained aging detection model after reaching a preset convergence condition.
2. The method for detecting the wet heat aging of an unsaturated polyester resin material according to claim 1, characterized in that: The step of obtaining the color feature vector of the unsaturated polyester resin material to be detected comprises: Acquire an RGB image of the unsaturated polyester resin material to be detected and convert it into a Lab color space to obtain a converted image; Based on the converted image, the global statistical features and color histogram features of each color channel are obtained to form a color feature vector of the unsaturated polyester resin material to be detected.
3. The method for detecting the wet heat aging of an unsaturated polyester resin material according to claim 1, characterized in that: The step of obtaining a training sample set comprises: Prepare multiple unsaturated polyester resin material samples with consistent size, shape and quality; Placing the prepared multiple unsaturated polyester resin material samples in an aging test box, applying different temperatures, humidity and aging time to perform wet heat aging treatment, and obtaining multiple unsaturated polyester resin material samples with different aging degrees; Acquire RGB images of unsaturated polyester resin material samples with different aging degrees and convert them into Lab color space to obtain multiple converted images; based on each converted image, obtain global statistical features and color histogram features of each color channel to form color feature vectors of unsaturated polyester resin material samples with different aging degrees; Based on multiple unsaturated polyester resin material samples with different aging degrees, an aging degree label corresponding to each color feature vector is obtained; wherein the aging degree label uses the relative change rate of the dielectric loss factor as the aging degree indicator, and the relative change rate The calculation expression is: ； In the formula, is the dielectric loss factor of the sample after aging treatment; is the dielectric loss factor of the sample without aging treatment.
4. A method for detecting wet heat aging of unsaturated polyester resin material according to claim 2 or 3, characterized in that: The global statistical features include mean value, standard deviation, skewness and kurtosis; the color histogram features include entropy, energy and contrast of the histogram.
5. The method for detecting the wet heat aging of an unsaturated polyester resin material according to claim 3, characterized in that: In the step of obtaining the aging degree label corresponding to each color feature vector based on a plurality of unsaturated polyester resin material samples with different aging degrees, the dielectric loss factor of the sample without aging treatment and the sample after aging treatment are measured and obtained at room temperature and 50 Hz.
6. The method for detecting the wet heat aging of an unsaturated polyester resin material according to claim 3, characterized in that: In the step of placing the prepared multiple unsaturated polyester resin material samples in an aging test box, applying different temperatures, humidity and aging times to perform wet heat aging treatment, and obtaining multiple unsaturated polyester resin material samples with different aging degrees, the humidity ranges from 50% to 95%, the temperature ranges from 50°C to 90°C, and the aging time ranges from 168h to 840h.
7. The method for detecting the wet heat aging of an unsaturated polyester resin material according to claim 1, characterized in that: The step of training the aging detection model using the training sample set, adjusting the model parameters, and obtaining the trained aging detection model after reaching the preset convergence condition comprises: Dividing the training sample set into a training set and a validation set; Based on the training set, the aging detection model is trained, model parameters are adjusted, and a trained aging detection model is obtained; Based on the validation set, a prediction performance evaluation is performed on the trained aging detection model, and when the evaluation result meets the preset performance requirement, a trained aging detection model is obtained; Among them, the indicators used in the prediction performance evaluation include mean absolute error, mean square error and determination coefficient.
8. A wet heat aging detection system for unsaturated polyester resin materials, characterized in that: include: A color feature vector acquisition module, used to acquire the color feature vector of the unsaturated polyester resin material to be detected; A prediction module is used to make predictions based on the acquired color feature vectors using a trained aging detection model to obtain an aging degree prediction value; A comparison module, used to compare the obtained aging degree prediction value with a preset aging degree threshold value to obtain an aging degree detection and classification result; in, The aging detection model adopts a support vector machine regression model, a random forest regression model or a multi-layer perceptron model; The training steps of the aging detection model include: Acquire a training sample set; wherein the training sample set includes color feature vectors of unsaturated polyester resin material samples of different aging degrees and aging degree labels corresponding to each color feature vector; the aging degree label corresponding to each color feature vector uses the relative change rate of the dielectric loss factor as an aging degree indicator; The aging detection model is trained using the training sample set, and the model parameters are adjusted to obtain a trained aging detection model after reaching a preset convergence condition.
9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the method for detecting wet heat aging of unsaturated polyester resin materials according to any one of claims 1 to 7 is implemented.
10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for detecting wet heat aging of an unsaturated polyester resin material according to any one of claims 1 to 7 is implemented.