Method for predicting changes in coating film, system for predicting changes in coating film, and application software.

Abstract

To provide a method for predicting changes in a coating film using an artificial intelligence model trained on machine learning, which analyzes the relationship between information about the coating film and / or information about the paints that make up the coating film and the results of weather resistance tests on the coating film, based on information about the coating film and / or information about the paints used to form the coating film. [Solution] A method for predicting changes in a coating film using an artificial intelligence model trained on the relationship between information about the coating film and / or information about the paint used to form the coating film and the results of weather resistance tests of the coating film.

Description

[Technical Field]

[0001] The present invention relates to a method for predicting changes in a coating film, a coating film change prediction system for predicting changes in a coating film, and application software. [Background technology]

[0002] In paint development, understanding how the paint film formed by applying the paint changes in the usage environment, particularly how it deteriorates, is one of the important technical considerations. Until now, paint development has been carried out by exposure tests in the usage environment and harsh environments, where painted panels were placed, and the results were taken into consideration. However, obtaining the results of exposure tests takes years, which has been one of the obstacles to rapid paint development. For this reason, accelerated weathering tests are conducted using accelerated weathering testers to change the coating film in a short time. However, there is often a discrepancy between the results of actual exposure tests and tests using accelerated weathering testers, and improvements have been needed.

[0003] Furthermore, exposure tests require the creation of a large number of painted panels for each type of paint, and there is also the task of transporting and placing the painted panels at the exposure testing site. Similarly, accelerated weathering tests using accelerated weathering testing machines also require the creation of a large number of painted panels for each type of paint, which is a significant burden in terms of cost and workload, and improvements were desired.

[0004] Studies have been conducted to date on predicting the deterioration of coatings. For example, Patent Documents 1 to 7 disclose methods or systems for predicting the deterioration of coatings. [Prior art documents] [Patent Documents]

[0005] [Patent Document 1] Japanese Patent Publication No. 2005-17132 [Patent Document 2] Japanese Patent Publication No. 2008-196887 [Patent Document 3] Japanese Patent Publication No. 2005-156309 [Patent Document 4] Japanese Patent Publication No. 2022-124933 [Patent Document 5] Japanese Patent Publication No. 2015-206603 [Patent Document 6] Japanese Patent Application Publication No. 57-197447 [Patent Document 7] Chinese Patent Application Publication No. 116976196 Specification [Overview of the Initiative] [Problems that the invention aims to solve]

[0006] The methods or systems disclosed in Patent Documents 1 to 7 for predicting the deterioration of coating films do not use an artificial intelligence model that has been trained on the relationship between information about the coating film and / or information about the paint constituting the coating film and the results of weathering tests of the coating film, and do not predict changes in the coating film from information about the coating film and / or information about the paint used to form the coating film. Therefore, the accuracy of predicting changes in the coating film was not satisfactory.

[0007] One of the problems that the present invention aims to solve is to provide a method for predicting changes in a coating film from information about the coating film and / or information about the paint used to form the coating film, using an artificial intelligence model that has been trained on the relationship between information about the coating film and / or information about the paint constituting the coating film and the results of weather resistance tests of the coating film. One of the problems that the present invention aims to solve is to provide a coating change prediction system that predicts changes in a coating from information about the coating and / or information about the paint used to form the coating, and which includes an artificial intelligence model that has been trained on the relationship between information about the coating and / or information about the paint constituting the coating and the results of weather resistance tests of the coating. One of the problems that the present invention aims to solve is to provide a method for predicting changes in a coating film and a coating film change prediction system that can predict the degree of change in the coating film over time with high accuracy. One of the problems that the present invention aims to solve is to provide application software for causing an information processing device to execute a method for predicting changes in the coating film, or application software for controlling and operating the coating film change prediction system. [Means for solving the problem]

[0008] As a result of diligent research to solve the above problems, the inventors have found that the above problems can be solved by (i) a method for predicting changes in a coating film using a specific artificial intelligence model, (ii) a coating film change prediction system equipped with a specific artificial intelligence for predicting changes in a coating film, (iii) application software for causing an information processing device to execute the method for predicting changes in a coating film, and (iv) application software for controlling and operating the system for predicting changes in a coating film, and have completed the present invention. In other words, the present invention is as described in the following sections 1 to 13.

[0009] [Section 1] A method for predicting changes in a paint film using an artificial intelligence model trained on the relationship between information about the paint film and / or information about the paint used to form the paint film and the results of weather resistance tests of the paint film. [Section 2] The method according to claim 1, wherein the change in the coating film includes one or more changes selected from the group consisting of color tone, water resistance, acid resistance, chemical resistance, surface roughness, gloss, shine, smoothness, adhesion, water contact angle, oil contact angle, scratch resistance, and appearance of the coating film. [Section 3] The method according to item 1 or 2, wherein the information on the coating film includes one or more selected from the group consisting of the composition of the coating film, information on the material to which the coating film is applied, the conditions for the formation of the coating film, the conditions for measuring the coating film, and the conditions for testing the coating film. [Section 4] The method according to any one of items 1 to 3, wherein the information of the coating film includes the structure of the coating film, and the structure of the coating film includes the layer structure of the coating film and the thickness of each layer. [Section 5] The method according to any one of items 1 to 4, wherein the information of the coating film includes the measurement conditions of the coating film, and the measurement conditions of the coating film include the presence or absence of polishing of the coating film surface and / or the presence or absence of water washing of the coating film surface when measuring the coating film. [Item 6] The method according to any one of items 1 to 5, wherein the information of the paint includes one or more selected from the group consisting of the type and blending amount of the resin for forming the coating film, the type and blending amount of the pigment, the type of the brightening material, the color tone and its blending amount, the type and blending amount of the additive, the type and blending amount of the pigment paste, and the type and blending amount of the solvent. [Item 7] The method according to any one of items 1 to 6, wherein the information of the paint includes the type and blending amount of the pigment paste. [Item 8] The method according to any one of items 1 to 7, wherein the coating film is a multi-layer coating film including a color base coating film and a clear coating film. [Item 9] The method according to any one of items 1 to 8, wherein the machine learning uses one or more of prediction algorithms by decision tree, linear regression, partial least squares regression, lasso regression, ridge regression, polynomial regression, Gaussian process regression, support vector machine, random forest, gradient boosting decision tree, K-nearest neighbor method, neural network, graph neural network, Bayesian estimation, or ensemble learning of these. [Item 10] The method according to any one of items 1 to 9, including predicting the weather resistance of automobile painting. [Item 11] A coating film change prediction system that predicts changes in a coating film from information on the coating film and / or information on the paint used for forming the coating film, and has an artificial intelligence model in which the relationship between the information on the coating film and / or the information on the paint used for forming the coating film and the results of the weather resistance test of the coating film is machine-learned. [Item 12] Application software for causing an information processing device to execute the method according to any one of items 1 to 10. [Item 13] Application software for controlling and operating the coating film change prediction system according to item 11.

Advantages of the Invention

[0010] According to the present invention, there is provided a method for predicting changes in a coating film from information on the coating film and / or information on the paint constituting the coating film, using an artificial intelligence model in which the relationship between the information on the coating film and / or the information on the paint and the results of a weather resistance test of the coating film is machine-learned. According to the present invention, there is provided a coating film change prediction system that predicts changes in a coating film from information on the coating film and / or information on the paint used for forming the coating film, and includes an artificial intelligence model in which the relationship between the information on the coating film and / or the information on the paint and the results of a weather resistance test of the coating film is machine-learned. According to the present invention, there are provided a method for predicting changes in a coating film and a coating film change prediction system for predicting changes in a coating film, which can predict the degree of change over time of the coating film with high accuracy. According to the present invention, there is provided application software for causing an information processing apparatus to execute the method for predicting changes in the coating film, or application software for controlling and operating the coating film change prediction system.

Brief Description of the Drawings

[0011] [Figure 1] Flow chart of a method for predicting changes in a coating film according to an embodiment of the present invention. [Figure 2] Diagram showing a configuration example of a coating film change prediction system according to an embodiment of the present invention.

Modes for Carrying Out the Invention

[0012] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. It should be understood that the present invention is not limited to the following embodiments, and includes various modifications implemented within the scope of not changing the gist of the present invention.

[0013] [Method for Predicting Changes in a Coating Film] The present invention provides a method for predicting changes in a coating film, which uses an artificial intelligence model trained on the relationship between information about the coating film and / or information about the paint used to form the coating film and the results of weathering tests of the coating film. By predicting changes in the coating film, it is possible to provide information on the deterioration and aging of the coating film, and the aging and evaluation of the appearance of painted articles and physical properties caused by the coating. The method for predicting changes in the coating film of the present invention is preferably carried out using a coating film change prediction system described later.

[0014] <coating film> In the method for predicting changes in a coating film according to the present invention, the "coating film" is not particularly limited as long as it is a coating film formed by applying paint to a substrate. The coating film may be a single-layer coating film consisting of only one layer, or a multi-layer coating film formed by applying two or more types of paint in layers. Examples of multi-layer coating films include those having two or more coating films, such as an electrodeposited coating film, an intermediate coating film, a color base coating film (color base coat layer), a glossy coating film, a colored coating film, and a clear coating film (clear coat layer). In the coating film change prediction method of the present invention, the coating film is preferably a multi-layer coating film (for example, (a) a multi-layer coating film including a color base coating film and a clear coating film, (b) a multi-layer coating film including an electrodeposited coating film, a color base coating film and a clear coating film, (c) a multi-layer coating film including an intermediate coating film, a color base coating film and a clear coating film, (d) a multi-layer coating film including an electrodeposited coating film, an intermediate coating film, a color base coating film and a clear coating film, etc.). When forming a coating on top of a coating, the coating may be formed on a wet coating, or on a coating that has been set and / or preheated and / or cured. When forming the coating film, each coating film may be heated and cured as it is formed, or multiple cured coating films may be formed simultaneously by heating any number of uncured coating films at the same time. In the present invention, multiple uncured coating films can be formed to create an uncured multilayer coating film, which is then heat-cured, thereby simultaneously heat-curing multiple uncured coating films.

[0015] <Changes in the coating> In the method for predicting changes in a coating film according to the present invention, the "changes in the coating film" to be predicted are changes in the coating film after an arbitrary amount of time has elapsed since immediately after the coating film was manufactured, and include at least changes in the appearance of the coating film and / or changes in the physical properties of the coating film.

[0016] Changes in the coating film include, for example, one or more changes selected from the group consisting of color tone, water resistance, acid resistance, chemical resistance, surface roughness, gloss, shine, smoothness, adhesion, contact angle, scratch resistance, electrical resistance, film thickness, stain resistance, solar reflectance, ultraviolet transmittance, haze, light transmittance, light reflectance, weather resistance, flip-flop properties, micro-glossiness, and coating film appearance. Of these, it is preferable that one or more changes selected from the group consisting of color tone, water resistance, acid resistance, chemical resistance, surface roughness, gloss, shine, smoothness, adhesion, water contact angle, oil contact angle, scratch resistance, and coating film appearance are included.

[0017] Changes in color tone include, for example, fading and whitening, L * a * b * Color system, XYZ color system, Munsell color system, L * u * v * Color systems, RGB color system, Yxy color system, Hunter Lab color system, L * C * The change in values ​​for each element in various color systems, such as the h color system, can be expressed as the change between the value immediately after coating application and the value of the coating after a specified period of time has elapsed. Alternatively, the change may be based on the measurement results of spectral reflectance or the evaluator's visual perception.

[0018] Changes in water resistance, acid resistance, chemical resistance, surface roughness, gloss, shine, smoothness, adhesion, contact angle, scratch resistance, electrical resistance, film thickness, stain resistance, solar reflectance, ultraviolet transmittance, light reflectance, weather resistance, flip-flop properties, and micro-glossiness can be expressed, for example, by using a physical property measuring device to show the change between the value immediately after coating application and the value of the coating after a specified period of time. Changes based on the evaluator's visual and tactile perception may also be used.

[0019] Changes in the appearance of the paint film include, for example, changes due to the occurrence of paint film cracks, blistering, peeling, scratches, contamination, shrinkage, and blistering, as well as changes in paint film gloss and sheen. The degree of these changes in the appearance of the paint film can be confirmed by the evaluator's visual inspection, and can also be quantified by means of instrumental measurement and image diagnosis.

[0020] <Information about the coating film> In the method for predicting changes in a coating film according to the present invention, "coating film information" includes one or more of the following: the composition of the coating film, information on the material to which the coating film is applied, the conditions for the formation of the coating film, the conditions for measuring the coating film, and the conditions for testing the coating film.

[0021] The composition of the coating film includes information about the coating film itself, such as one or more of the following: the layer structure of the coating film (single layer, multi-layer, number of layers, layering order, function of each layer, etc.), coating film thickness (color base coating film thickness, clear coating film thickness), coating film color tone, coating film properties (reflectance, light transmittance, water absorption rate, contact angle, gloss, surface roughness, micro-glossiness, flip-flop properties, water repellency, water-slip properties, smoothness, haze, opacity, etc.), coating film shape, coating film application, and painted product part number (product number, paint code, etc.). In particular, from the viewpoint of improving prediction accuracy, it is preferable that the composition of the coating film includes the layer structure of the coating film and the coating film thickness.

[0022] Information about the material to be coated includes, for example, one or more of the following: the material of the material to be coated, the physical properties of the material, the surface condition of the material, the surface treatment of the material, the shape of the material, and the use of the material (use of the painted object (part name, area, etc.)).

[0023] Examples of conditions for the formation of a coating include one or more of the following: the painting method, painting work conditions (orientation of the painted surface on the object to be painted (vertical surface, inclined surface, flat surface, etc.), temperature during painting, humidity during painting, painting location, person who applies the coating, etc.), the painting equipment used, the operating conditions of the painting equipment, and drying conditions (baking temperature, maximum temperature reached, heating time, heating rate, cooling rate, cooling time).

[0024] Measurement conditions for the coating film include, for example, one or more of the following: measuring equipment, measurement method, measurement conditions, measurer, and pretreatment during measurement (such as whether the coating surface is polished or washed with water when measuring the exposed plate after exposure). In particular, from the viewpoint of improving prediction accuracy, it is preferable to include information on the pretreatment of the coating film during measurement, especially information on whether the coating surface is polished and / or whether the coating surface is washed with water.

[0025] Examples of test conditions for coatings include one or more test conditions in exposure tests (exposure location, exposure period, exposure timing, etc.) and test conditions in accelerated tests (type of accelerated test, accelerated test equipment, accelerated test conditions, etc.).

[0026] <Information on paints used for film formation> In the method for predicting changes in a coating film according to the present invention, "information on the paint used for coating film formation" includes, for example, one or more of the following: information on the paint itself, information on the manufacturing conditions of the paint, information on each component that makes up the paint / information on each component used in the manufacture of the paint.

[0027] Information about the paint itself may include, for example, the paint's composition, color, physical properties (gloss, viscosity, storage modulus, loss tangent, electrical resistance, contact angle, solar reflectance, UV transmittance, weather resistance, viscoelasticity, stringiness, haze, light transmittance, foaming properties, surface tension, solids content, pH, storage stability, repellency, circulation properties, etc.), paint product number / product name, and paint supplier, etc.

[0028] Information related to the manufacturing conditions of the paint may include one or more items such as the paint manufacturer, the time of manufacture, the place of manufacture, and the manufacturing conditions of the paint (dispersion conditions, temperature conditions, equipment used, etc.).

[0029] Information on each component that makes up the paint / each component used in the manufacture of the paint includes, for example, information on each component contained in the paint and each component used in the manufacture of the paint. For example, this includes one or more pieces of information such as the type, shape (e.g., form, aspect ratio, thickness, primary particle diameter, secondary particle diameter, particle diameter distribution, etc.), properties (e.g., curability, thermoplasticity, solubility, dispersibility, dispersed particle diameter, molecular weight, molecular weight distribution, heat resistance temperature, glass transition temperature, softening point, reactivity, etc.) and their blending amount, the type, shape, properties (e.g., hardness, surface properties, presence or absence of surface modification, wettability, dispersibility, pH, zeta potential, etc.), color tone and their blending amount, the type, shape, properties, color tone and their blending amount of the glossing agent, the type, shape, properties and their blending amount of the dispersant, the type, properties and their blending amount of the solvent, the type, shape, properties and their blending amount of the additives, the type, properties, composition and their blending amount of the premix such as pigment paste, the manufacturing method, manufacturing date, manufacturer, seller, product number, lot number, etc. In particular, from the viewpoint of improving prediction accuracy, it is preferable to include information on the premix, especially information on the type of pigment paste and / or its amount.

[0030] <Results of weather resistance test of coating film> In the method for predicting changes in a coating film according to the present invention, "results of the coating film weathering test" refers to data obtained from weathering tests of the coating film, such as outdoor exposure, indoor exposure, and weathering test results using an accelerated weathering tester (e.g., xenon weather meter, sunshine weather meter, ultraviolet fluorescent lamp weather meter, ultraviolet fade meter). The data may include measurements taken with measuring instruments, evaluation results from panelists or other evaluators, etc. The types of data obtained from the weathering resistance test of the coating film include one or more data related to the changes in the coating film, and may also include data related to the period of the weathering resistance test.

[0031] <Machine-learned artificial intelligence model> In the method for predicting changes in a coating film according to the present invention, the algorithm used for machine learning when creating the "machine-learned artificial intelligence model" is not particularly limited as long as it is a machine learning algorithm capable of creating an artificial intelligence model. For example, one or more prediction algorithms such as decision trees, linear regression, partial least squares regression, Lasso regression, ridge regression, polynomial regression, Gaussian process regression, support vector machines, random forests, gradient boosting decision trees, K-nearest neighbors, neural networks, graph neural networks, Bayesian estimation, or ensemble learning of these may be used. In this invention, it is preferable to use one or more of the following as machine learning algorithms: random forests and neural networks, and it is more preferable to use an algorithm based on random forests.

[0032] When performing machine learning on artificial intelligence models, appropriate machine learning frameworks and libraries can be used depending on the data scale, execution environment, etc. These frameworks and libraries may be developed by the user, or one or more of the available ones may be used. Examples of machine learning frameworks and libraries include TensorFlow, PyTorch, Keras, Caffe, Microsoft Cognitive Toolkit, MXNet, Chainer, Scikit-learn, OpenCV, and spaCy.

[0033] Machine learning may be either supervised or unsupervised. In the present invention, it is preferable to use supervised learning, which involves associating information about the coating film and / or information about the paint used to form the coating film with the results of weather resistance tests of the coating film.

[0034] For machine learning, information on the coating film and / or information on the paint used to form the coating film, as well as the results of weather resistance tests of the coating film, can be input by transmitting data via wired, wireless, or a combination thereof, or via a recording medium. Examples of input using communication means include one or more combinations of various communication networks such as LAN (Local Area Network), WAN (Wide Area Network), the Internet, and telephone networks. Input via a recording medium can be performed by reading data from a recording medium such as a magnetic recording medium, optical recording medium, or paper recording medium using an appropriate reading means.

[0035] In machine learning, the reliability of information on coating films and / or paints used to form them, as well as the results of weathering tests on coating films, can be enhanced by improving consistency (consistency of labeling criteria), reproducibility (highly reproducible labeling for the same data), and transparency (clarity of the data creation process and the basis for labeling).

[0036] In machine learning, it is preferable to preprocess raw data related to information on coating films and / or paints used to form the coating films, as well as the results of weather resistance tests of the coating films, to create training data. Examples of preprocessing include processing such as imputing missing data, filtering such as removing outliers, data normalization / standardization, and arithmetic processing (logarithmic scaling, power scaling, etc.). This makes the data applicable to various machine learning algorithms, and also allows for data specialized for specific machine learning algorithms.

[0037] <Prediction of changes in the coating film> Figure 1 is a flowchart of a method for predicting changes in a coating film according to one embodiment of the present invention. A method for predicting changes in a coating film according to one embodiment of the present invention includes at least the following steps S01 to S05: Step S01: Obtain information on the coating film and / or information on the paint used to form the coating film, as well as the results of the weather resistance test of the coating film. Step S02: Machine learning is used to determine the relationship between information about the coating film and / or information about the paint used to form the coating film and the results of the weather resistance test of the coating film. Step S03: Creating an artificial intelligence model Step S04: Input of information about the coating film and / or information about the paint used to form the coating film. Step S05: Use an artificial intelligence model to predict changes in the coating. It includes.

[0038] (Step S01) Step S01 is a step in which information on the coating film and / or information on the paint used to form the coating film, as well as the results of the weathering test of the coating film, are obtained. Details of the information on the coating film, the information on the paint used to form the coating film, and the results of the weathering test of the coating film are as described in <Information on the coating film>, <Information on the paint used to form the coating film>, and <Results of the weathering test of the coating film> above, respectively. In Step S01, the information on the coating film and / or information on the paint used to form the coating film and the corresponding results of the weathering test of the coating film are obtained in a manner that is related to each other.

[0039] Information on the coating film and / or information on the paint used to form the coating film, as well as the results of weathering tests on the coating film, may be obtained by importing them into a database within a system that predicts changes in the coating film, or by accessing an external database via a communication line.

[0040] Information on the coating film, information on the paint used to form the coating film, and the results of the weather resistance test of the coating film can be easily obtained in a linked manner by forming the coating film, conducting a weather resistance test on the coating film, recording the results of the weather resistance test, and accumulating this data in a database. This allows information on the coating film and / or information on the paint used to form the coating film to be easily obtained in a linked manner with the corresponding results of the weather resistance test of the coating film.

[0041] (Step S02) Step S02 is a step in which machine learning is performed on the relationship between information about the coating film and / or information about the paint used to form the coating film and the results of the weather resistance test of the coating film. The algorithm for machine learning is as described in the <machine-learned artificial intelligence model> above. By associating information about the coating film and / or information about the paint used to form the coating film with the results of weather resistance tests of the coating film, and then using machine learning to perform this process, supervised machine learning can be achieved.

[0042] Machine learning can be initiated by inputting information about the coating film and / or the paint used to form the coating film, as well as the results of weather resistance tests on the coating film. During input, the information about the coating film and / or the paint used to form the coating film, as well as the results of weather resistance tests on the coating film, are entered using an appropriate data input method. This input may also be performed by transferring data from a database or importing (downloading) data.

[0043] (Step S03) Step S03 is the step for creating an artificial intelligence model. In step S02, by using machine learning to study the relationship between information about the coating film and / or information about the paint used to form the coating film and the results of weathering tests on the coating film, it is possible to create an artificial intelligence model that takes information about the coating film and / or the paint used to form the coating film as input and outputs predictions of changes in the coating film.

[0044] (Step S04) Step S04 is the step of inputting information about the coating film and / or information about the paint used to form the coating film into the artificial intelligence model. The information about the coating film and / or the paint used to form the coating film that is input may be different from the information used during machine learning, or it may be the same information.

[0045] By inputting information about a coating film and / or paint used for coating formation that differs from the information used in machine learning, it is possible to predict changes in the coating film at any given time for a desired coating film or a coating film obtained from a desired paint. This makes it possible to provide information on the deterioration and aging of the coating film, as well as the aging and evaluation of the appearance of painted articles and the physical properties caused by the coating.

[0046] By inputting the same coating information and / or paint information used for coating formation that was used in machine learning, the performance and versatility of the artificial intelligence model created by machine learning can be verified and evaluated. Furthermore, it is possible to predict future changes in the coating and to predict changes in the coating at a point in time when specific measurements were not taken.

[0047] The method for inputting information about the coating film and / or the paint used to form the coating film into the artificial intelligence model is not particularly limited. Similar to the information input during machine learning in step S02, information can be input using an appropriate data input means. Information input may also be performed by extracting, transferring, and importing (downloading) information from a database. Information input can also be performed using a program.

[0048] (Step S05) Step S05 is a step in which an artificial intelligence model is used to predict changes in the coating film by calculating predictive data of changes in the coating film at any given time, based on the coating film information and / or paint information used to form the coating film that was input in step S04. When calculating predictive data for changes in the coating film at any given point in time, it is possible to calculate only one predictive data set, or to calculate multiple predictive data sets as a group of candidates. If multiple predictive data sets are calculated as a group of candidates, the optimal predictive data can be selected from the group of candidates based on predetermined criteria. These predetermined criteria may include, for example, selecting data sets that fall within a predetermined range by focusing on specific physical properties, or using numerical values ​​calculated from predetermined formulas.

[0049] When predicting changes in a coating film, by simultaneously predicting various physical properties and the like involved in the changes in the coating film, it is possible to perform highly accurate prediction taking into account the mutual influences. For example, by simultaneously predicting not only the changes in the appearance of the coating film but also the changes in the coating film physical properties, highly accurate prediction taking into account these mutual influences becomes possible. In the present invention, it is possible to simultaneously predict two or more of the changes in the appearance of the coating film, and it is possible to simultaneously predict two or more of the physical properties of the coating film. For example, ΔE and L * a * b * It is possible to simultaneously predict each value of, and it is also possible to predict any one of these individually. When predicting changes in a coating film, by using the prediction data output from an artificial intelligence model as a variation amount, it is possible to cancel out factors such as variations and measurement errors, thereby enhancing the accuracy of the prediction.

[0050] When predicting changes in a coating film, by programming steps S01 to S05, particularly steps S04 to S05, it is possible to predict the changes in the coating film as a group even for coating films in which various components, conditions, etc. included in the coating film information and various components, conditions, etc. included in the coating formation paint information are varied within a certain range. Also, it is possible to confirm whether the input is appropriate, and further, it is possible to confirm the compatibility of system updates and the like.

[0051] (Other steps) In the method for predicting changes in a coating film according to an embodiment of the present invention, in addition to the above steps S01 to S05, it is possible to include one or more other steps. Other steps include, for example, a data preprocessing step prior to step S01, which preprocesses raw data related to information on the coating film and / or information on the paint used to form the coating film, as well as the results of weathering tests on the coating film; an artificial intelligence model verification and evaluation step following step S03, which verifies and evaluates the created artificial intelligence model; and a prediction data selection step following step S05, which selects the optimal prediction data when multiple candidate data are calculated in step S05.

[0052] <Applications of methods for predicting changes in coating films> The method for predicting changes in a coating film according to the present invention can predict the state of the coating film at a specific point in the future, and therefore can be suitably used in applications where it is necessary to predict changes in the coating film on a painted object. For example, it can be used to predict the durability of paint on products, predict changes in paint films on buildings and structures in the civil engineering and construction industries (such as predicting maintenance such as repainting timing), and promote the commercial value of paints. In particular, it can predict the weather resistance of paint on products such as automobiles, which is useful when developing and proposing paints.

[0053] [A system that predicts changes in the coating film.] The coating film change prediction system of the present invention is a system that predicts changes in a coating film from information about the coating film and / or information about the paint used to form the coating film, and includes an artificial intelligence model that has been trained on the relationship between information about the coating film and / or information about the paint used to form the coating film and the results of weather resistance tests of the coating film. The coating film change prediction system of the present invention is suitably used in the method for predicting changes in a coating film according to the present invention.

[0054] Figure 2 shows an example of the configuration of a coating film change prediction system 1 according to one embodiment of the present invention. The coating change prediction system 1 shown in Figure 2 is equipped with a computer 10, and the coating change prediction system 1 shown in Figure 2 is configured within the computer 10. The coating change prediction system 1 comprises an artificial intelligence model 101, a coating change prediction unit 102, an (internal) database 103a, a reception unit 104, a display unit 105, and a recording unit 106. The coating change prediction system 1 is connectable to user terminals 20a to 20c of users 2a to 2c via a communication line N, and is also connectable to an (external) database 103b via the communication line N. Furthermore, the display unit 105 is connectable to a virtual space M, and is configured to display images of coating changes in the virtual space M.

[0055] Computer 10 is under the control of the operator or provider of the coating change prediction system 1. The operator or provider of the coating change prediction system 1 is the provider of the method for predicting changes in the coating of the present invention, or a person commissioned or entrusted by the provider. Users 2a to 2c are not particularly limited as long as they are people who use the coating change prediction system 1. For example, one or more may be paint users, paint manufacturers, painting contractors, painted product manufacturers, etc. User terminals 2a to 2c are not particularly limited as long as they are terminals that can connect to the coating change prediction system 1 via the communication line N. For example, one or more may be laptop computers, mobile phones (smartphones), personal digital assistants, etc. Although only users 2a to 2c are shown in Figure 2, the number of users is not particularly limited. The operator or provider of the coating change prediction system 1 may also charge at least some users for the use of the system or for some operations performed while using the system.

[0056] The coating change prediction system 1 obtains information on coatings and / or paints used to form the coatings, as well as the results of weathering tests on the coatings, from information sources such as the (internal) database 103a, the (external) database 103b, and external measuring instruments (not shown), sometimes via a communication line N, and uses this information to train the artificial intelligence model 101. When obtaining information on coatings and / or paints used to form the coatings, as well as the results of weathering tests on the coatings, from sources other than the (internal) database 103a, such as the (external) database 103b or external measuring instruments, and performing machine learning on the artificial intelligence model 101, the system can do so via the reception unit 104. In this case, data verification and preprocessing can be performed at the reception unit 104.

[0057] After the machine learning of the artificial intelligence model 101 is completed, information about the coating film and / or information about the paint used to form the coating film is input via an input means (not shown) on the computer 10, or via one or more input means from user terminals 20a to 20c via the communication line N. Based on the input information, the coating film change prediction system 1 uses the artificial intelligence model 101 to predict the changes in the coating film in the coating film change prediction unit 102 and calculates the result.

[0058] The calculated prediction of changes in the coating can be displayed in the display unit 105 and / or in the virtual space M connected to the display unit 105. Furthermore, the calculated prediction of changes in the coating can be displayed on user terminals 20a to 20c via the communication line N, and users can participate in the virtual space M via the user terminals 20a to 20c. The calculated prediction of changes in the coating film can be stored in the recording unit 106, and the prediction accuracy of the coating film change prediction system 1 can be improved by comparing it with the measured values ​​of changes in the coating film as needed.

[0059] [Application Software] The application software of the present invention is application software that causes an information processing device to execute a method for predicting changes in the coating film, and functions to execute the method for predicting changes in the coating film of the present invention. The application software of the present invention is application software for controlling and operating the coating film change prediction system, and functions to control and operate the coating film change prediction system of the present invention.

[0060] The application software of the present invention is an information processing terminal that performs the method for predicting changes in the coating film of the present invention, a coating film change prediction system of the present invention, or equipment that performs each means constituting the information processing terminal or system, etc., and has an HDD (Hard Disk Drive) or flash memory. It may be pre-stored in recording devices such as mori. Alternatively, it may be installed on equipment using wireless or wired communication means, or removable recording media such as DVDs, CD-ROMs, or USB memory sticks. [Explanation of symbols]

[0061] 1. Coating film change prediction system 10 Computers 101 Artificial Intelligence Models 102 Coating film change prediction section 103a (Internal) Database 103b (External) Databases 104 Reception Department 105 Display section 106 Records Section N communication line M Virtual Space 2a~2c users 20a~20c User-side terminal

Claims

1. A method for predicting changes in a paint film using an artificial intelligence model trained on the relationship between information about the paint film and / or information about the paint used to form the paint film and the results of weather resistance tests of the paint film.

2. The method according to claim 1, wherein the change in the coating film includes one or more changes selected from the group consisting of color tone, water resistance, acid resistance, chemical resistance, surface roughness, gloss, shine, smoothness, adhesion, water contact angle, oil contact angle, scratch resistance, and appearance of the coating film.

3. The method according to claim 1 or 2, wherein the information on the coating film includes one or more selected from the group consisting of the composition of the coating film, information on the material to which the coating film is applied, the conditions for the formation of the coating film, the conditions for measuring the coating film, and the conditions for testing the coating film.

4. The method according to claim 1 or 2, wherein the information of the coating film includes the structure of the coating film, and the structure of the coating film includes the layer structure of the coating film and the thickness of each layer.

5. The method according to claim 1 or 2, wherein the information of the coating film includes measurement conditions for the coating film, and the measurement conditions for the coating film include, when measuring the coating film, whether or not the surface of the coating film is polished and / or whether or not the surface of the coating film is washed with water.

6. The method according to claim 1 or 2, wherein the information on the paint includes one or more selected from the group consisting of the type and amount of film-forming resin, the type and amount of pigment, the type of glossing agent, the color tone and amount thereof, the type and amount of additive, the type and amount of pigment paste, and the type and amount of solvent.

7. The method according to claim 1 or 2, wherein the information on the paint includes the type of pigment paste and the amount thereof.

8. The method according to claim 1 or 2, wherein the coating film is a multilayer coating film including a color base coating film and a clear coating film.

9. The method according to claim 1 or 2, wherein the machine learning method uses one or more of the following prediction algorithms: decision trees, linear regression, partial least squares regression, Lasso regression, ridge regression, polynomial regression, Gaussian process regression, support vector machines, random forests, gradient boosting decision trees, K-nearest neighbors, neural networks, graph neural networks, Bayesian estimation, or ensemble learning thereof.

10. The method according to claim 1 or 2, including the prediction of the weather resistance of automobile paint.

11. A paint film change prediction system that predicts changes in the paint film from information about the paint film and / or information about the paint used to form the paint film, and which includes an artificial intelligence model trained on the relationship between information about the paint film and / or information about the paint used to form the paint film and the results of weather resistance tests of the paint film.

12. Application software for causing an information processing device to execute the method described in claim 1 or 2.

13. Application software for controlling and operating the coating film change prediction system described in claim 11.

Images