Yellowing detection method of polyvinyl acetal based on machine learning

Abstract

The present invention discloses a polyvinyl acetal yellowing detection method based on machine learning, which relates to the technical field of polymer material detection. The detection method includes an experimental modeling stage and a model application stage. In the experimental modeling stage, relative absorbance is designed as an aging evaluation standard to obtain a relative absorbance time curve; a first mapping relationship is fitted using the relative absorbance time curve and the first yellowing degree measured experimentally; a neural network model is trained and constructed according to the relative absorbance and mechanical property parameters; in the application stage, a second yellowing degree of the actual material to be tested is obtained, and several actual relative absorbances are obtained according to the second yellowing degree and several first mapping relationships, and the several actual relative absorbances are input into the neural network model to obtain the actual mechanical property parameters. Through the experimental modeling stage and the model application stage, the present invention can more accurately estimate the mechanical properties of the polyvinyl acetal material in the actual aging process according to the actual yellowing degree.

Description

Technical Field

[0001] The present invention relates to the technical field of polymer material detection, and in particular to a method for detecting yellowing of polyvinyl acetal based on machine learning. Background Art

[0002] Polyvinyl Acetal (PVAc) is a polymer formed by the reaction of polyvinyl alcohol (PVA) and aldehydes (usually formaldehyde). It is a thermoplastic material with good transparency, excellent adhesion and water resistance. Due to its excellent physical and chemical properties, it is widely used in many fields, such as for the manufacture of various coatings and adhesives, food packaging and protective films, manufacturing optical lenses and other optical components; used as electronic component packaging materials to provide protection and insulation; in the pharmaceutical industry, PVA can be used to prepare drug release carriers and control the release rate of drugs, etc. It not only plays an important role in industrial applications, but also shows good prospects in high-tech fields such as medicine and optics.

[0003] Polyvinyl acetal (PVAc) is affected by many factors during use, including light, temperature, humidity, oxygen, etc., and thus undergoes a series of complex aging processes. The aging process of polyvinyl acetal involves complex chemical and physical changes, including free radical formation, cross-linking, degradation, etc. During the aging process of polyvinyl acetal, certain functional groups in the molecular chain (such as aldehyde groups) may undergo oxidation reactions, causing the color of the material to turn yellow. Yellowing not only affects the appearance of the material, but may also be related to changes in the molecular structure, reflecting the degree of degradation and cross-linking within the material. As aging proceeds, the mechanical properties of polyvinyl acetal (such as tensile strength, toughness, impact resistance, etc.) will be affected. The combined effects of factors such as free radical generation, cross-linking, degradation, and changes in crystallinity lead to a decrease in the strength and toughness of the material, which manifests as increased brittleness.

[0004] Therefore, in practical applications, understanding and controlling the aging process of polyvinyl acetal is crucial to ensure its performance and service life. Existing technology is an effective method to reflect the aging process of polyvinyl acetal (PVAc) by the degree of yellowing, because yellowing is usually closely related to the chemical structure change, degradation and aging mechanism of the material. As the degree of yellowing increases, the mechanical properties of polyvinyl acetal (such as tensile strength, toughness, etc.) usually decrease, but it is difficult to estimate the mechanical properties of polyvinyl acetal during the aging process based on the degree of yellowing in the existing technology.

[0005] There is an urgent need for a yellowing detection method for polyvinyl acetal, which can estimate the mechanical properties of polyvinyl acetal during the aging process based on the detected yellowing degree. Summary of the invention

[0006] The present invention provides a polyvinyl acetal yellowing detection method based on machine learning to solve at least one of the problems mentioned in the above background technology.

[0007] The specific technical solutions provided by this application are as follows:

[0008] The polyvinyl acetal yellowing detection method based on machine learning includes the following steps:

[0009] S1. Perform an aging test on a polyvinyl acetal sample and construct a relative absorbance time curve;

[0010] S2. Obtaining the first yellowing degree and mechanical property parameters of the polyvinyl acetal samples corresponding to a certain aging time;

[0011] S3, constructing a first mapping relationship between the first yellowing degree and the relative absorbance according to the relative absorbance time curve and the first yellowing degree corresponding to the aging time;

[0012] S4, steps S1-S3 are repeated n times to obtain n first mapping relationships and n mechanical characteristic parameter groups; each mechanical characteristic parameter group represents a number of mechanical characteristic parameters corresponding to a number of aging times in one cycle;

[0013] S5, setting a training parameter set according to the relative absorbance and mechanical property parameter group, and constructing and training a neural network model according to the training parameter set;

[0014] S6. Obtain a second yellowing degree of the actual polyvinyl acetal material, and compare the second yellowing degree with the n first mapping relationships to obtain n actual relative absorbances of the actual polyvinyl acetal material, and input the n actual relative absorbances into the trained neural network model to obtain actual mechanical property parameters.

[0015] Preferably, the relative absorbance time curve is a curve of the relationship between relative absorbance and aging time; and the mechanical property parameters include impact strength, tensile strength and bending strength.

[0016] Preferably, the step S1 specifically includes:

[0017] Step S11, setting the aging time, and selecting several aging time points;

[0018] Step S12, performing an aging test: placing the polyvinyl acetal sample in a constant temperature and humidity aging box, and performing aging treatment according to a set aging time;

[0019] Step S13: sampling the first absorbance of the first infrared light wave i and the second absorbance of the second infrared light wave j at a certain aging time, and recording the first absorbance Ai and the second absorbance Aj ;

[0020] Step S14: Calculate relative absorbance , the calculation formula is as follows:

[0021] , that is, relative absorbance Indicates the first absorbance Ai and the second absorbance Aj ratio.

[0022] Step S15: constructing a relative absorbance time curve μij ( t ), specifically:

[0023] ,

[0024] In the formula, Ai ( t ) is the time t The first absorbance at Aj ( t ) is the time t The second absorbance at .

[0025] Preferably, a relative absorbance time curve is constructed μij ( t ) also includes:

[0026] Step S16: each relative absorbance time curve μij ( t ) and the corresponding wave number and Associative storage, where the wave number refers to the wave number of the first infrared light wave i; wave number Refers to the wave number of the second infrared light wave j.

[0027] Preferably, in step S2, obtaining the first yellowing degree of the polyvinyl acetal sample corresponding to a plurality of aging times includes:

[0028] Step S21, obtain the original CIE Lab value, and record the original CIE Lab value as ( , , ),in, represents the original brightness, Represents the original color components from green to red, Represents the original color components from blue to yellow;

[0029] Step S22: During the aging test, at each aging time point, a spectrophotometer is used to measure the current CIE Lab value and record it as ( , , ),in, Indicates the current brightness. Represents the current color component from green to red, Indicates the current color component from blue to yellow;

[0030] Step S23, calculating the first yellowing degree of the polyvinyl acetal sample, comprising:

[0031] Calculate the brightness change value : = - ;

[0032] Calculate the yellowness change value : = - ;

[0033] Obtain the two-dimensional vector of yellowing degree: Yellowing degree = ( , ).

[0034] Preferably, in step S2, the mechanical property parameters of the polyvinyl acetal samples corresponding to several aging times are obtained, including:

[0035] Step S201, preparing polyvinyl acetal samples corresponding to a certain aging time according to standard requirements to obtain polyvinyl acetal film samples corresponding to a certain aging time that meet the specifications;

[0036] Step S202, using an impact tester to measure the impact strength of the polyvinyl acetal film samples corresponding to a number of aging times, and recording the results;

[0037] Step S203, using a universal testing machine to test the tensile strength of the polyvinyl acetal film samples corresponding to a number of aging times, and recording the results;

[0038] Step S204, using a universal testing machine to test the bending strength of the polyvinyl acetal film samples corresponding to a number of aging times, and recording the results;

[0039] Step S205 , comparing the test results of steps S203 - S204 with the mechanical property parameters of a standard polyvinyl acetal film, to obtain the effects of different aging times on the mechanical properties of the polyvinyl acetal film.

[0040] Preferably, step S2 further includes:

[0041] Step S300: The yellowing degree corresponding to different aging times ( , ) and mechanical property parameters are organized into a table; the influence of aging time on the yellowing degree and mechanical property parameters is analyzed through the table, and the trend and correlation are observed.

[0042] Preferably, obtaining the second yellowing degree of the actual polyvinyl acetal material in step S6 includes:

[0043] Step S61, obtaining the original CIE Lab value ( , , );

[0044] Step S62: Use a spectrophotometer to measure the current CIE Lab value and record it as ( , , ),in, Indicates the current brightness. Represents the current color component from green to red, Indicates the current color component from blue to yellow;

[0045] Step S63, calculating the second yellowing degree of the actual polyvinyl acetal material, comprising:

[0046] Calculate the brightness change value : = - ;

[0047] Calculate the yellowness change value : = - ;

[0048] Obtain the two-dimensional vector of yellowing degree: Yellowing degree = ( , ).

[0049] Preferably, obtaining the actual second yellowing degree of the polyvinyl acetal material in step S6 further includes:

[0050] Step S64, judging whether the actual polyvinyl acetal material turns yellow, specifically:

[0051] :

[0052] like <0, it means the brightness of the material is reduced.

[0053] like >0, it means the brightness of the material is improved;

[0054] :

[0055] like >0, it means the color of the material turns yellow.

[0056] like <0, it means the color of the material turns blue. The smaller the value, the more obvious the blue change.

[0057] Preferably, in step S6, the second yellowing degree is compared with n first mapping relationships to obtain n actual relative absorbances of the actual polyvinyl acetal material, including: comparing step S4 to obtain n first mapping relationships, obtaining the actual time point of the actually measured second yellowing degree in the first mapping relationships, and obtaining n actual relative absorbances in the n first mapping relationships according to the actual time point.

[0058] Compared with the prior art, the present invention has the following beneficial effects:

[0059] The present invention effectively constructs a relative absorbance time curve of polyvinyl acetal material, thereby evaluating its performance changes under different aging conditions, which provides important reference data for the long-term application of polyvinyl acetal. At the same time, the relative absorbance is obtained by the ratio method, which has high sensitivity and accuracy, and reduces the influence of factors such as the measurement environment on a single absorbance value.

[0060] The present invention can evaluate the performance change of polyvinyl acetal material in long-term use by obtaining yellowing degree and mechanical property parameters corresponding to several aging times and performing comprehensive analysis thereof, thus providing important reference data for the application of the material.

[0061] The present invention establishes a mapping relationship between the degree of yellowing and the relative absorbance by analyzing the relative absorbance time curve and the corresponding aging time. This process links the optical properties with the material properties, provides a basis for subsequent analysis, and can better understand how the degree of yellowing affects the relative absorbance, thereby enhancing the ability to predict material properties.

[0062] The above-mentioned multiple sets of mapping relationships and mechanical property parameters of the present invention provide rich training data, which is helpful to improve the generalization ability of the neural network model. At the same time, through different light wave numbers, the performance of the material under different conditions can be more comprehensively evaluated.

[0063] The neural network model completed through training in the present invention can accurately predict corresponding mechanical property parameters according to new relative absorbance data.

[0064] The present invention obtains the actual relative absorbance corresponding to the second yellowing degree from the first mapping relationship between the yellowing degree and the relative absorbance constructed in the experimental modeling stage according to the second yellowing degree of the polyvinyl acetal material actually measured in the model application stage, takes the actual relative absorbance as the input of the trained neural network model, and outputs the actual mechanical property parameters of the polyvinyl acetal material through the model. Thus, the mechanical properties of the polyvinyl acetal material in the actual aging process can be more accurately estimated through the actual yellowing degree. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0066] 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 or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0067] Figure 1 A schematic flow chart of a polyvinyl acetal yellowing detection method based on machine learning provided in an embodiment of the present invention;

[0068] Figure 2 A schematic flow chart of step S1 in the polyvinyl acetal yellowing detection method based on machine learning provided in an embodiment of the present invention;

[0069] Figure 3 A schematic flow chart of step S2 in the method for detecting yellowing of polyvinyl acetal based on machine learning provided in an embodiment of the present invention;

[0070] Figure 4 A schematic flow chart of step S6 in the polyvinyl acetal yellowing detection method based on machine learning provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0071] 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.

[0072] It should be noted that all directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present invention are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication will also change accordingly.

[0073] In addition, the descriptions of "first", "second", etc. in the present invention are only used for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least the feature. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0074] Yellowing is the phenomenon of a material changing color during aging, usually from clear or colorless to yellow or brown. This change is usually related to changes in the chemical structure of the polymer.

[0075] The yellowing of polyvinyl acetal during the aging process mainly includes oxidation reaction, photodegradation and hydrolysis reaction. Oxidation reaction refers to the oxidation reaction that may occur in polyvinyl acetal under the influence of light, heat or oxygen, resulting in molecular chain breakage and cross-linking, thereby causing yellowing. Photodegradation refers to the breakage and cross-linking of polymer chains caused by ultraviolet light irradiation, the formation of pigments, and thus yellowing. Hydrolysis reaction refers to the hydrolysis of PVAc in a humid environment, resulting in changes in its structure and color.

[0076] The degree of yellowing can usually be quantitatively measured using tools such as UV-Vis spectroscopy, a colorimeter or a colorimeter. Commonly used indicators include the yellowness index (YI) and the color difference value (ΔE).

[0077] Machine learning is a branch of artificial intelligence that aims to automatically improve the performance of models through data and experience. It uses algorithms and statistical models to enable computers to learn from data and make predictions or decisions without being explicitly programmed. As technology continues to develop, machine learning will play an increasingly important role in the field of materials science. Machine learning provides new perspectives and tools for the detection of yellowing of polyvinyl acetal. By integrating multiple data sources, establishing predictive models, and realizing real-time monitoring, machine learning can not only help understand the aging mechanism of materials, but also optimize the performance and application of materials, providing efficient and accurate data analysis and prediction methods for the detection of yellowing of polyvinyl acetal.

[0078] To this end, the present invention provides a polyvinyl acetal yellowing detection method and system based on machine learning, and the specific implementation process of the present invention will be described in detail below.

[0079] Embodiment 1

[0080] See also Figure 1 The present invention provides a polyvinyl acetal yellowing detection method based on machine learning, and the detection method includes an experimental modeling stage and a model application stage.

[0081] In the experimental modeling stage, relative absorbance is designed as an aging evaluation standard to obtain a relative absorbance time curve; a first mapping relationship is fitted using the relative absorbance time curve and the first yellowing degree measured experimentally; a neural network model is trained and constructed based on relative absorbance and mechanical property parameters;

[0082] In the model application stage, the second yellowing degree of the actual polyvinyl acetal material to be tested is obtained, a number of actual relative absorbances are obtained according to the second yellowing degree and a number of first mapping relationships, and the number of actual relative absorbances are input into the neural network model to obtain actual mechanical property parameters;

[0083] The present invention can more accurately estimate the mechanical properties of the polyvinyl acetal material in the actual aging process according to the actual yellowing degree through the experimental modeling stage and the model application stage.

[0084] Specifically, the experimental modeling phase includes the following steps:

[0085] S1. Perform an aging test on the polyvinyl acetal sample and construct a relative absorbance time curve. The relative absorbance time curve is a curve showing the relationship between relative absorbance (y-axis) and aging time (x-axis); the aging test is performed in a constant temperature and humidity aging chamber, and the polyvinyl acetal sample is subjected to thermal oxygen aging treatment. The thermal oxygen aging process of the polyvinyl acetal is characterized from the perspective of chemical structure changes by three characterization methods: nuclear magnetic resonance hydrogen spectrum (H-NMR), Fourier transform infrared spectrum (FTIR) and ultraviolet-visible absorption spectrum (UV-Vis). The specific aging test process is carried out by conventional means in the art, and the present invention is not limited thereto.

[0086] In an embodiment of the present invention, Figure 2 As shown, an aging test is performed on the polyvinyl acetal sample to construct a relative absorbance time curve, specifically including:

[0087] Step S11, set the aging time, and select several aging time points, such as 0 hours, 5 hours, 10 hours, 15 hours, etc.

[0088] Step S12, performing an aging test: placing the polyvinyl acetal sample in a constant temperature and humidity aging box, and performing aging treatment according to a set aging time.

[0089] Step S13: sampling the first absorbance of the first infrared light wave i and the second absorbance of the second infrared light wave j at a certain aging time, and recording the first absorbance Ai and the second absorbance Aj .

[0090] Step S14: Calculate relative absorbance , the calculation formula is as follows:

[0091] , that is, relative absorbance Indicates the first absorbance Ai and the second absorbance Aj ratio.

[0092] Step S15: constructing a relative absorbance time curve μij ( t ), specifically:

[0093] ,

[0094] In the formula, Ai ( t ) is the time t The first absorbance at Aj ( t ) is the time t The second absorbance at .

[0095] Specifically, the relative absorbance at different time points μ The data is organized into tables based on aging time. t is the horizontal axis (x-axis), relative absorbance μ As the ordinate (y-axis), draw the relative absorbance time curve μij ( t ).

[0096] In a preferred embodiment of the present invention, a relative absorbance time curve is constructed. μij ( t ) also includes:

[0097] Step S16: each relative absorbance time curve μij ( t ) and the corresponding wave number and Perform associative storage for subsequent analysis and query.

[0098] Among them, the wave number refers to the wave number of the first infrared light wave i; wave number Refers to the wave number of the second infrared light wave j.

[0099] The wave number of the first infrared light wave i An infrared spectrum characteristic peak corresponding to the heat-sensitive group of polyvinyl acetal;

[0100] The wave number of the second infrared light wave An infrared spectrum characteristic peak corresponding to the non-thermosensitive group of polyvinyl acetal.

[0101] Among them, thermosensitive groups refer to chemical groups that are sensitive to thermal changes within a specific temperature range. These groups may undergo changes in chemical structure under the action of heat, resulting in changes in the physical, chemical or optical properties of the material. They usually show characteristic absorption peaks in the infrared spectrum, which can be used to identify and analyze the thermal sensitivity of materials. In polyvinyl acetal (PVAc), common thermosensitive groups include aldehyde groups (-CHO) and alcohol groups (-OH). These groups may react at a certain temperature and affect the performance of the material. For example:

[0102] Aldehyde groups may polymerize or cross-link when heated.

[0103] Alcohol groups may lose water under certain conditions, causing structural changes in the material.

[0104] The present invention can effectively construct a relative absorbance time curve of polyvinyl acetal material through the above steps, and then evaluate its performance changes under different aging conditions. This provides important reference data for the long-term application of polyvinyl acetal. At the same time, the relative absorbance is obtained by the ratio method, which has high sensitivity and accuracy, and reduces the influence of factors such as the measurement environment on a single absorbance value.

[0105] Example: In a specific embodiment of the present invention, taking polyvinyl butyral as an example, for example, the wave number of the first infrared light wave in step S1 is 3465, 1732, 1114 or 996 ; The wave number of the second infrared light wave is 1378 ;

[0106] Among them, 3465 and 1732 The infrared spectrum characteristic peaks at 1114 correspond to the heat-sensitive groups hydroxyl and carbonyl of polyvinyl butyral. and 996 The infrared spectrum characteristic peak of the acetal ring of the thermosensitive group of polyvinyl butyral is 1378 The infrared spectrum characteristic peak at corresponds to the non-heat-sensitive group methyl of polyvinyl butyral.

[0107] Preferably, for example, in one cycle, the wave number of the first infrared light wave in step S1 is 1732 , the wave number of the second infrared light wave is 1378 ; The relative absorbance time curve It is expressed as:

[0108] .

[0109] in, represents the ratio of the first absorbance to the second absorbance; t represents the aging time, and is the fitting parameter, .

[0110] Wave number The absorbance of light waves (A1732) corresponds to the stretching vibration of the carbonyl group (C=O). In the early stage of aging, the ester group in polyvinyl acetal decomposes, and the generated ketones and aldehydes react, and the carbonyl content decreases rapidly; in the later stage of aging, the molecular chains are mainly cross-linked, and the carbonyl content of the generated ketones and aldehydes decreases slowly.

[0111] Wave number The absorbance of light waves (A1378) corresponds to the symmetrical bending vibration of the methyl group (CH3). The methyl group is relatively stable in the structure of PVB and is not prone to significant changes. Therefore, A1378 can be used as a stable reference point.

[0112] The A1732 / A1378 ratio reflects the relative abundance of carbonyl (C=O) groups relative to methyl (CH 3 ) groups to visually assess the aging degree of the material.

[0113] S2. Obtaining the first yellowing degree and mechanical property parameters of the polyvinyl acetal samples corresponding to a certain aging time.

[0114] The first yellowing degree is obtained by obtaining the CIE Lab value of the polyvinyl acetal sample through a spectrophotometer and calculating; the CIE Lab color space consists of three components:

[0115] : Lightness, ranging from 0 (black) to 100 (white);

[0116] : Color components from green to red;

[0117] : Color components from blue to yellow.

[0118] Among them, the mechanical property parameters include impact strength, tensile strength and bending strength.

[0119] Specifically, Figure 3 As shown, the first yellowing degree of the polyvinyl acetal sample corresponding to several aging times is obtained, including:

[0120] Step S21, obtaining the original CIE Lab value, specifically using a spectrophotometer to measure on an unaged polyvinyl acetal sample, and recording the original CIE Lab value ( , , ),in, represents the original brightness, Represents the original color components from green to red, Represents the original color components from blue to yellow;

[0121] Step S22: During the aging test, at each aging time point, a spectrophotometer is used to measure the current CIE Lab value and record it as ( , , ),in, Indicates the current brightness. Represents the current color component from green to red, Indicates the current color component from blue to yellow;

[0122] Step S23, calculating the first yellowing degree of the polyvinyl acetal sample, comprising:

[0123] Calculate the brightness change value : = - ;

[0124] Calculate the yellowness change value : = - ;

[0125] Obtain the two-dimensional vector of yellowing degree: Yellowing degree = ( , ).

[0126] Step S24, yellowing determination of the polyvinyl acetal sample, specifically:

[0127] :

[0128] like <0, it means the brightness of the material is reduced.

[0129] like >0, it means the brightness of the material is improved;

[0130] :

[0131] like >0, it means the color of the material turns yellow (turns toward yellow).

[0132] like <0, it means the color of the material turns blue. The smaller the value, the more obvious the blue change.

[0133] Specifically, mechanical property parameters of polyvinyl acetal samples corresponding to several aging times are obtained, including:

[0134] Step S201, preparing polyvinyl acetal samples corresponding to a certain aging time according to standard requirements to obtain polyvinyl acetal film samples corresponding to a certain aging time that meet the specifications;

[0135] Step S202, using an impact tester to measure the impact strength of the polyvinyl acetal film samples corresponding to a number of aging times, and recording the results;

[0136] Step S203: using a universal testing machine to test the tensile strength of the polyvinyl acetal film samples corresponding to a number of aging times, and recording the results.

[0137] Step S204: Use a universal testing machine to test the bending strength of polyvinyl acetal film samples corresponding to several aging times, and record the results. Step S205: Compare the test results of steps S203-S204 with the mechanical property parameters of the standard polyvinyl acetal film to obtain the effects of different aging times on the mechanical properties of the polyvinyl acetal film.

[0138] In a preferred embodiment of the present invention, obtaining the first yellowing degree and mechanical property parameters of the polyvinyl acetal samples corresponding to a plurality of aging times also includes:

[0139] Step S300: The yellowing degree corresponding to different aging times ( , ) and mechanical properties (impact strength, tensile strength, flexural strength) are organized into a table; the influence of aging time on the yellowing degree and mechanical properties is analyzed through the table, and the trend and correlation are observed.

[0140] The present invention can evaluate the performance change of polyvinyl acetal material in long-term use by obtaining yellowing degree and mechanical property parameters corresponding to several aging times and performing comprehensive analysis thereof, thus providing important reference data for the application of the material.

[0141] S3, constructing a first mapping relationship between the first yellowing degree and the relative absorbance according to the relative absorbance time curve and the first yellowing degree corresponding to the aging time;

[0142] The present invention establishes a mapping relationship between the degree of yellowing and the relative absorbance by analyzing the relative absorbance time curve and the corresponding aging time. This process links the optical properties with the material properties, provides a basis for subsequent analysis, and can better understand how the degree of yellowing affects the relative absorbance, thereby enhancing the ability to predict material properties.

[0143] S4, repeating the process n times: reselecting the wave number of the first infrared light wave and / or the second infrared light wave, and then executing steps S1-S3; after executing the process n times, obtaining n first mapping relationships and n mechanical characteristic parameter groups; each mechanical characteristic parameter group represents a number of mechanical characteristic parameters corresponding to a number of aging times in one cycle;

[0144] Step S4 selects different infrared light wave numbers and repeats steps S1-S3 to obtain multiple sets of data, which helps to capture the impact of different light waves on the optical properties of the material. At the same time, each cycle generates a new set of mapping relationships and mechanical property parameter groups, increasing the richness and reliability of the data.

[0145] The above-mentioned multiple sets of mapping relationships and mechanical property parameters of the present invention provide rich training data, which is helpful to improve the generalization ability of the neural network model. At the same time, through different light wave numbers, the performance of the material under different conditions can be more comprehensively evaluated.

[0146] S5, model construction and training: setting a training parameter set according to the relative absorbance and mechanical property parameter group, and constructing and training a neural network model according to the training parameter set; the training parameter set includes several groups of relative absorbance and mechanical property parameters obtained under the same aging time; the input of the neural network model is n relative absorbances, and the output is the mechanical property parameters.

[0147] Among them, the neural network model can adopt the neural network models commonly used in the field, such as feedforward neural network (FNN), convolutional neural network (CNN), recurrent neural network (RNN), etc.; the construction and training process of the neural network model can adopt the conventional methods in the field, and the present invention is not limited here.

[0148] For several groups of relative absorbances obtained under the same aging time, each group of relative absorbances obtained under the same aging time includes n relative absorbances, that is, the aging time corresponding to these n relative absorbances is the same, and the difference lies in the wave number of the corresponding first infrared light wave i and / or the wave number of the second infrared light wave j Different. Since the mechanical property parameters obtained for the same material at the same aging time are almost the same, for the n mechanical parameter groups obtained in step S4, after removing the abnormal values, the average value of the n mechanical property parameter groups at each aging time is taken as the mechanical property parameter corresponding to the n relative absorbance at the aging time. Based on the above content, after determining the mechanical property parameters corresponding to the n relative absorbance at each aging time, the n relative absorbance at each aging time and its corresponding mechanical property parameters are added to the training parameter set to train the neural network model to fit the relationship between the n relative absorbances and the mechanical property parameters.

[0149] The neural network model trained by the above steps can accurately predict the corresponding mechanical property parameters according to the new relative absorbance data. The output of the model can provide a scientific basis for the selection and application of materials, helping engineers and researchers make better decisions.

[0150] Specifically, the model application phase includes the following steps:

[0151] S6. Obtain a second yellowing degree of the actual polyvinyl acetal material, and compare the second yellowing degree with the n first mapping relationships to obtain n actual relative absorbances of the actual polyvinyl acetal material, and input the n actual relative absorbances into the trained neural network model to obtain actual mechanical property parameters.

[0152] Among them, Figure 4 As shown, obtaining the second yellowing degree of the actual polyvinyl acetal material includes:

[0153] Step S61, obtaining the original CIE Lab value ( , , );

[0154] Step S62: Use a spectrophotometer to measure the current CIE Lab value and record it as ( , , ),in, Indicates the current brightness. Represents the current color component from green to red, Indicates the current color component from blue to yellow;

[0155] Step S63, calculating the second yellowing degree of the actual polyvinyl acetal material, comprising:

[0156] Calculate the brightness change value : = - ;

[0157] Calculate the yellowness change value : = - ;

[0158] Obtain the two-dimensional vector of yellowing degree: Yellowing degree = ( , ).

[0159] Step S64, judging whether the actual polyvinyl acetal material turns yellow, specifically:

[0160] :

[0161] like <0, it means the brightness of the material is reduced.

[0162] like >0, it means the brightness of the material is improved;

[0163] :

[0164] like >0, it means the color of the material turns yellow (turns toward yellow).

[0165] like <0, it means the color of the material turns blue. The smaller the value, the more obvious the blue change.

[0166] Among them, according to the second yellowing degree, comparing the n first mapping relationships, obtaining n actual relative absorbances of the actual polyvinyl acetal material includes: comparing step S4 to obtain n first mapping relationships, obtaining the actual time point of the actually measured second yellowing degree in the first mapping relationship, and obtaining n actual relative absorbances in the n first mapping relationships according to the actual time point.

[0167] The present invention obtains the actual relative absorbance corresponding to the second yellowing degree from the first mapping relationship between the yellowing degree and the relative absorbance constructed in the experimental modeling stage according to the second yellowing degree of the polyvinyl acetal material actually measured in the model application stage, takes the actual relative absorbance as the input of the trained neural network model, and outputs the actual mechanical property parameters of the polyvinyl acetal material through the model. Thus, the mechanical properties of the polyvinyl acetal material in the actual aging process can be more accurately estimated through the actual yellowing degree.

[0168] The foregoing is merely a specific embodiment of the present invention, which enables those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A method for detecting yellowing of polyvinyl acetal based on machine learning, characterized in that: The steps include: S1. Perform an aging test on a polyvinyl acetal sample and construct a relative absorbance time curve; S2. Obtaining the first yellowing degree and mechanical property parameters of the polyvinyl acetal samples corresponding to a certain aging time; S3, constructing a first mapping relationship between the first yellowing degree and the relative absorbance according to the relative absorbance time curve and the first yellowing degree corresponding to the aging time; S4, steps S1-S3 are repeated n times to obtain n first mapping relationships and n mechanical characteristic parameter groups; each mechanical characteristic parameter group represents a number of mechanical characteristic parameters corresponding to a number of aging times in one cycle; S5, setting a training parameter set according to the relative absorbance and mechanical property parameter group, and constructing and training a neural network model according to the training parameter set; S6, obtaining a second yellowing degree of the actual polyvinyl acetal material, and obtaining n actual relative absorbances of the actual polyvinyl acetal material according to the second yellowing degree and comparing the n first mapping relationships, and inputting the n actual relative absorbances into the trained neural network model to obtain actual mechanical property parameters; In step S2, obtaining the first yellowing degree of the polyvinyl acetal sample corresponding to a plurality of aging times includes: Step S21, obtain the original CIE Lab value, and record the original CIE Lab value as ( L 1 , a 1 , b 1 ),in, L 1 represents the original brightness, a 1 Represents the original color components from green to red, b 1 Represents the original color components from blue to yellow; Step S22: During the aging test, at each aging time point, a spectrophotometer is used to measure the current CIE Lab value and record it as ( L 2 , a 2 , b 2 ),in, L 2 Indicates the current brightness. a 2 Represents the current color component from green to red, b 2 Indicates the current color component from blue to yellow; Step S23, calculating the first yellowing degree of the polyvinyl acetal sample, comprising: Calculate the brightness change ΔL :ΔL = L 2 - L 1 ; Calculate the yellowness change value Δb :Δb = b 2 - b 1 ; Obtain the two-dimensional vector of yellowing degree: Yellowing degree = (Δ L ,Δ b ); In step S2, mechanical property parameters of polyvinyl acetal samples corresponding to several aging times are obtained, including: Step S201, preparing polyvinyl acetal samples corresponding to a certain aging time according to standard requirements to obtain polyvinyl acetal film samples corresponding to a certain aging time that meet the specifications; Step S202, using an impact tester to measure the impact strength of the polyvinyl acetal film samples corresponding to a number of aging times, and recording the results; Step S203, using a universal testing machine to test the tensile strength of the polyvinyl acetal film samples corresponding to a number of aging times, and recording the results; Step S204, using a universal testing machine to test the bending strength of the polyvinyl acetal film samples corresponding to a number of aging times, and recording the results; Step S205, comparing the test results of steps S203-S204 with the mechanical property parameters of a standard polyvinyl acetal film to obtain the effects of different aging times on the mechanical properties of the polyvinyl acetal film; In step S6, obtaining the actual second yellowing degree of the polyvinyl acetal material includes: Step S61, obtaining the original CIE Lab value ( L 1 , a 1 , b 1 ); Step S62: Use a spectrophotometer to measure the current CIE Lab value and record it as ( L 3 , a 3 , b 3 ),in, L 3 Indicates the current brightness. a 3 Represents the current color component from green to red, b 3 Indicates the current color component from blue to yellow; Step S63, calculating the second yellowing degree of the actual polyvinyl acetal material, comprising: Calculate the brightness change ΔL ':ΔL '= L 3 - L 1 ; Calculate the yellowness change value Δb ':Δb '= b 3 - b 1 ; Obtain the two-dimensional vector of yellowing degree: Yellowing degree = (Δ L ',Δ b '); Step S64, judging whether the actual polyvinyl acetal material turns yellow, specifically: D L ': If Δ L '<0, it means the brightness of the material is reduced. If Δ L '>0, it means the brightness of the material is improved; D b ': If Δ b '>0, it means the color of the material turns yellow, If Δ b '<0, it means the color of the material turns blue. The smaller the value, the more obvious the blue change.

2. The method for detecting yellowing of polyvinyl acetal based on machine learning according to claim 1, characterized in that: The relative absorbance time curve is a curve of the relationship between relative absorbance and aging time; the mechanical property parameters include impact strength, tensile strength and bending strength.

3. The method for detecting yellowing of polyvinyl acetal based on machine learning according to claim 1, characterized in that: The step S1 specifically includes: Step S11, setting the aging time, and selecting several aging time points; Step S12, performing an aging test: placing the polyvinyl acetal sample in a constant temperature and humidity aging box, and performing aging treatment according to a set aging time; Step S13: sampling the first absorbance of the first infrared light wave i and the second absorbance of the second infrared light wave j at a certain aging time, and recording the first absorbance Ai and the second absorbance Aj ; Step S14: Calculate relative absorbance , the calculation formula is as follows: , that is, relative absorbance Indicates the first absorbance Ai and the second absorbance Aj The ratio of Step S15: constructing a relative absorbance time curve μij ( t ), specifically: ， In the formula, Ai ( t ) is the time t The first absorbance at Aj ( t ) is the time t The second absorbance at .

4. The method for detecting yellowing of polyvinyl acetal based on machine learning according to claim 3, characterized in that: Constructing relative absorbance time curves μij ( t ) also includes: Step S16: each relative absorbance time curve μij ( t ) and the corresponding wave number and Associative storage, where the wave number refers to the wave number of the first infrared light wave i; wave number Refers to the wave number of the second infrared light wave j.

5. The method for detecting yellowing of polyvinyl acetal based on machine learning according to claim 1, characterized in that: Step S2 also includes: Step S300: The yellowing degree (ΔL , Δb ) and mechanical property parameters are organized into a table; the influence of aging time on the yellowing degree and mechanical property parameters is analyzed through the table, and the trend and correlation are observed.

6. The method for detecting yellowing of polyvinyl acetal based on machine learning according to claim 1, characterized in that: In step S6, the second yellowing degree is compared with the n first mapping relationships to obtain n actual relative absorbances of the actual polyvinyl acetal material, including: comparing step S4 to obtain n first mapping relationships, obtaining the actual time point of the actually measured second yellowing degree in the first mapping relationships, and obtaining n actual relative absorbances in the n first mapping relationships according to the actual time point.

Images