Formula optimization method of two-component solvent-free brush coating type transparent waterproof polyurea

Abstract

The invention provides a method for optimizing a double-component solvent-free brush coating type transparent waterproof polyurea formula, and belongs to the technical field of polyurea formulae, the method for optimizing the double-component solvent-free brush coating type transparent waterproof polyurea formula comprises the following steps: collecting data of a plurality of swimming pool operation environments; obtaining multi-mode fingerprint data of each raw material in the polyurea formula; the method comprises the following steps: establishing a swimming pool use environment parameter model according to swimming pool operation environment data, and selecting a plurality of formulas with optimal swimming pool waterproof performance indexes by utilizing a swimming pool waterproof material quantitative structure-activity relationship chemical model and a pre-trained swimming pool special waterproof performance prediction model to generate performance vectors; performing an accelerated aging test in a laboratory simulation swimming pool environment; selecting a formula closest to a performance vector required by the preset two-component solvent-free brush coating type transparent waterproof polyurea for the swimming pool; and carrying out field test in a real swimming pool environment by adopting the optimized formula to obtain a final formula.

Description

Technical Field

[0001] The invention belongs to the technical field of polyurea formulations, and in particular relates to a method for optimizing a two-component solvent-free brush-coated transparent waterproof polyurea formulation. Background Art

[0002] As an indispensable public entertainment facility in modern cities, swimming pools have become an important place for people's daily entertainment and fitness. In the use of swimming pools, the performance of waterproof materials is particularly critical. On the one hand, environmental parameters such as temperature, pressure, and water quality of the swimming pool water will have a significant impact on waterproof materials. If the material performance cannot meet these environmental conditions, leakage, bulging, peeling and other problems are very likely to occur, threatening the safe use of swimming pools. On the other hand, as an outdoor public facility, swimming pools require waterproof materials to have good weather resistance and aging resistance to ensure that the materials maintain stable waterproof performance during long-term use. In addition, swimming pools also require waterproof materials to be transparent to meet visual aesthetic requirements, and also require anti-slip, easy to construct and other characteristics to create a comfortable and safe swimming experience for users.

[0003] At present, the common swimming pool waterproof materials on the market are mainly the following: polyvinyl chloride (PVC) waterproof membrane, epoxy resin waterproof coating, polymer cement-based waterproof coating, etc. These materials do meet the waterproof needs of swimming pools to a certain extent, but they also have many shortcomings. Although PVC waterproof membrane has good waterproof performance, it is not transparent, difficult to construct, and has environmental pollution problems; epoxy resin waterproof coating has excellent waterproof performance and weather resistance, but has strict requirements on the construction environment and has certain toxicity; although polymer cement-based waterproof coating is easy to use and has good construction performance, its waterproofness, weather resistance and anti-slip properties need to be further improved.

[0004] Therefore, developing a new waterproof material that can simultaneously meet all the performance requirements of swimming pools has become a key technical issue that needs to be solved urgently. The material must not only meet the standard requirements in terms of basic properties such as waterproofness, weather resistance and transparency, but also have good mechanical properties, construction adaptability and swimming pool-specific properties to ensure long-term and stable use. Summary of the invention

[0005] In view of this, the present invention provides a two-component solvent-free brush-coated transparent waterproof polyurea formula optimization method, which solves the problem that the existing polyurea formula needs to be verified by a large number of experiments and manual experience to obtain a formula suitable for swimming pools, which is time-consuming and labor-intensive.

[0006] The present invention is achieved in that: The present invention provides a method for optimizing the formulation of a two-component solvent-free brush-coated transparent waterproof polyurea, which comprises the following steps: S10, collecting operating environment data of multiple swimming pools, including water temperature change data, water pressure data, water quality parameter data, and ultraviolet irradiation intensity data; obtaining multimodal fingerprint data of each raw material in a two-component solvent-free brush-coated transparent waterproof polyurea formula for swimming pools; S20, establishing a swimming pool use environment parameter model according to the swimming pool operation environment data, for each raw material, selecting a material whose multimodal fingerprint data similarity with the raw material is greater than a preset fingerprint similarity threshold and meets the requirements of the swimming pool use environment parameter model as a substitute raw material, and generating a recipe set by using a traversal loop method; S30, using a quantitative structure-activity relationship chemical model for swimming pool waterproof materials and combining it with the swimming pool use environment parameter model, calculating the raw material ratio of each formula in the formula set; S40, using a pre-trained swimming pool-specific waterproof performance prediction model, inputting a multimodal fingerprint and a raw material ratio of each formula in the formula set, and obtaining a swimming pool waterproof performance index of each formula; S50, selecting multiple formulas with the best swimming pool waterproof performance index to form a first formula set, and obtaining multiple necessary properties of the polyurea material obtained by conducting a small test on each formula in the first formula set, and generating a performance vector; S60, performing an accelerated aging test in a laboratory simulating a swimming pool environment to test the aging resistance of each formulation in the first formulation set, and updating the performance vector; S70, according to the performance vector of each formula, selecting the formula closest to the required performance vector of the preset two-component solvent-free brush-coated transparent waterproof polyurea for swimming pools as the optimized formula; S80, using the optimized formula to conduct field tests in a real swimming pool environment, and recording material performance change data; S90. According to the material performance change data and in combination with the swimming pool use environment parameter model, the raw material ratio of the optimized formula is adjusted to obtain a final formula.

[0007] For step S10, the specific methods of collecting environmental data and acquiring fingerprint data are as follows: 1. Water temperature change data collection model: ; In the formula, is the water temperature at time t, in °C; is the average water temperature in °C; is the temperature fluctuation amplitude, in °C; is the time in hours; is a random fluctuation term, which follows a normal distribution .

[0008] 2. Water pressure data calculation model: ; In the formula, is the water pressure at depth h, in Pa; is the density of water, which is 1000kg / m³; is the acceleration due to gravity, with a value of 9.81m / s²; is the water depth in m; is the atmospheric pressure, which is 101325Pa.

[0009] 3. Multimodal fingerprint similarity calculation method: ; In the formula, is the overall similarity index; For the Similarity index of the modalities; is the weight coefficient of the i-th mode and satisfies ; is the total number of modes.

[0010] Similarity calculation for specific modalities: Spectral data similarity: ; In the formula, is the spectral similarity; For the two sets of spectral data to be compared, Absorption value at each wavelength point; are their respective average values; is the number of wavelength points.

[0011] Similarity of morphological data: ; In the formula, is the morphological similarity; The two images to be compared are at positions The gray value at ; is the size of the image; is the number of gray levels.

[0012] Thermal analysis data similarity: ; In the formula, is the thermal analysis similarity; For the two groups of samples to be compared, Thermal analysis data of temperature points; is the attenuation coefficient; is the number of temperature points.

[0013] For step S20, the recipe set generation method is: 1. Calculation of environmental parameter matching: ; In the formula, For the The matching degree of environmental parameters of the alternative raw materials; For the A matching function for environmental parameters; For the Alternative raw materials Environmental parameter values; is the total number of environmental parameters.

[0014] 2. Raw material substitution conditions: ; In the formula, For the Multimodal fingerprint similarity of alternative raw materials; is the preset fingerprint similarity threshold; is the matching degree of environmental parameters; Matches the threshold value for the preset environmental parameters.

[0015] For step S30, the raw material ratio calculation method is: 1. Quantitative structure-activity relationship model: ; In the formula, is the target performance indicator; is the ratio of each raw material; are model parameters; The quantity of raw materials.

[0016] 2. Environmental parameter constraints: ; ; In the formula, Respectively The lower and upper limits of the ratio of the raw materials.

[0017] For step S40, the performance prediction model: 1. Multimodal feature extraction: Spectral characteristics: ; Morphological features: ; Thermal analysis characteristics: ; 2. Feature Fusion: ; In the formula, is the fused feature vector; is the attention mechanism function.

[0018] 3. Performance index calculation: ; In the formula, is the waterproof performance index; For fusion features; is the raw material ratio vector; A multi-layer perceptron.

[0019] For step S50, the performance vector is constructed: ; In the formula, is the performance vector; to They are the standardized values ​​of waterproof performance, mechanical properties, weather resistance, construction performance, and special performance respectively.

[0020] For step S60, aging resistance evaluation: ; In the formula, is the performance value at time t; is the initial performance value; is the aging rate constant; is the aging time.

[0021] For step S70, the recipe selection criteria are: ; In the formula, is the performance vector distance; is the actual performance value; is the target performance value; is the weight coefficient.

[0022] For step S90, recipe optimization method: ; In the formula, For the optimized ratio; For the original ratio; is the adjustment factor; is the performance change rate.

[0023] On the basis of the above technical solution, the two-component solvent-free brush-coated transparent waterproof polyurea formulation optimization method of the present invention can also be improved as follows: The two-component solvent-free brush-coated transparent waterproof polyurea formula for swimming pools comprises component A and component B, wherein component A comprises aliphatic isocyanate prepolymer and additives, wherein the additives comprise anti-ultraviolet agent, anti-settling agent, defoaming agent, and curing accelerator, and component B comprises polyetheramine, amine chain extender and additives, wherein the additives comprise antioxidant, wetting dispersant, leveling agent, and toughening agent, and the mass ratio of component A to component B is 1:1 to 1:1.2.

[0024] Furthermore, the component A comprises 80 to 95 parts of the aliphatic isocyanate prepolymer and 5 to 20 parts of the auxiliary agent by weight, wherein the auxiliary agent comprises 1 to 3 parts of the anti-ultraviolet agent, 0.5 to 2 parts of the anti-settling agent, 0.3 to 1 part of the defoaming agent, and 0.5 to 2 parts of the curing accelerator; the component B comprises 60 to 80 parts of the polyether amine, 15 to 25 parts of the amine chain extender, and 5 to 15 parts of the auxiliary agent by weight, wherein the auxiliary agent comprises 1 to 2 parts of the antioxidant, 0.5 to 1.5 parts of the wetting and dispersing agent, 0.3 to 1 part of the leveling agent, and 2 to 5 parts of the toughening agent.

[0025] Furthermore, the multimodal fingerprint data refers to characteristic data of raw materials obtained by various analytical methods, including spectral data, morphological data, thermal analysis data and other characterization data; The spectral data includes one or more of infrared spectrum, Raman spectrum, ultraviolet spectrum and visible spectrum, which are used to reflect the existence and strength of chemical bonds; The morphology data includes one or more of a scanning electron microscope image, an atomic force microscope image, a transmission electron microscope image, and an optical microscope image, which are used to reflect the microscopic morphology and size distribution of the raw material; The thermal analysis data includes one or more of thermogravimetric analysis data, differential scanning calorimetry analysis data, thermomechanical analysis data, and dynamic mechanical thermal analysis data, which are used to reflect the thermal stability and phase change behavior of the raw materials; The other characterization data include one or more of X-ray diffraction data, particle size analysis data, specific surface area data, and density data, which are used to reflect the structure, composition and particle size information of the raw material.

[0026] Furthermore, the performance vector is a vector composed of waterproof performance, mechanical performance, weather resistance performance, construction performance, and swimming pool-specific performance; The waterproof performance includes one or more of the following: anti-seepage strength, water absorption rate, water permeability, and waterproof sealing performance; The mechanical properties include one or more of tensile strength, elongation at break, tear strength, and underwater adhesion; The weather resistance includes one or more of chloride ion corrosion resistance, ozone aging resistance, hydrolysis resistance, and temperature cycle resistance; The construction performance includes one or more of surface drying time, curing time, coating thickness, and underwater construction adaptability; The swimming pool-specific properties include one or more of resistance to disinfectants, algae resistance, slip resistance, and transparency.

[0027] Furthermore, the structure of the swimming pool waterproof performance prediction model is a multi-level performance prediction network structure, including five parallel performance prediction modules. Each module realizes information sharing and feature fusion through an interactive learning mechanism. The performance prediction modules include a basic waterproof performance prediction module, a mechanical performance prediction module, an environmental adaptability prediction module, a construction performance prediction module and a special performance prediction module.

[0028] Furthermore, the basic waterproof performance prediction module mainly predicts the basic waterproof performance index of the material; The mechanical property prediction module predicts the mechanical property index of the material in combination with the basic waterproof performance characteristics; The environmental adaptability prediction module integrates waterproof performance and mechanical performance characteristics to predict the weather resistance performance index of the material; The construction performance prediction module comprehensively considers the aforementioned performance characteristics to predict the construction application performance index of the material; The dedicated performance prediction module integrates all performance characteristics to predict the dedicated performance index of the material in the specific environment of the swimming pool.

[0029] Furthermore, each prediction module includes five network modules: spectral data prediction network module, morphological data prediction network module, thermal analysis data prediction network module, other characterization data prediction network module and multimodal fusion network module.

[0030] Furthermore, the spectral data prediction network module is used to extract the characteristics of the spectral data using a convolutional neural network or a recurrent neural network structure, and predict the performance index corresponding to the characteristics of the spectral data; The morphology data prediction network module is used to extract the morphology features of the morphology data using a convolutional neural network, and predict the performance indicators corresponding to the morphology features; The thermal analysis data prediction network module is used to extract thermal analysis features of thermal analysis data using a fully connected neural network or a long short-term memory network structure, and predict performance indicators corresponding to the thermal analysis features; The other characterization data prediction network module is used to extract other characterization features of other characterization data using a fully connected neural network or a long short-term memory network structure, and predict performance indicators corresponding to the other characterization features; The multimodal fusion network module is used to adopt an attention mechanism or a gating mechanism to fuse the different spectral data prediction network modules, morphological data prediction network modules, thermal analysis data prediction network modules, and other characterization data prediction network modules, and finally predict all performance indicators corresponding to all multimodal fingerprint data.

[0031] Furthermore, the water quality parameter data includes one or more of residual chlorine content, pH value, turbidity, and microbial content in water.

[0032] Compared with the prior art, the two-component solvent-free brush-coated transparent waterproof polyurea formula optimization method provided by the present invention has the following beneficial effects: First, the present invention fully considers various factors of the swimming pool use environment, including water temperature, water pressure, water quality, ultraviolet radiation, etc., and collects these environmental parameter data to establish a swimming pool use environment parameter model. At the same time, multimodal characteristic fingerprint data of various raw materials are also obtained, including multiple characterization methods such as spectrum, morphology, and thermal analysis. Through the environmental parameter model and raw material fingerprint similarity analysis, alternative raw materials that can meet the requirements of the swimming pool use environment are screened out, laying the foundation for formula optimization; Secondly, the present invention establishes a quantitative structure-activity relationship model for swimming pool waterproof materials, which quantitatively describes the mathematical relationship between the ratio of various raw materials in the formula and the final performance. Combined with the aforementioned environmental parameter model, the optimal raw material ratio of each alternative formula is obtained through optimization calculation. In addition, a multi-level performance prediction network is used to comprehensively consider multiple modal characteristics to accurately predict the comprehensive performance indicators of each formula in the swimming pool environment, including waterproofness, mechanical properties, weather resistance, construction performance, and specificity; Thirdly, the present invention conducted a small test and accelerated aging test on the candidate formula in a laboratory simulated swimming pool environment, and obtained the measured data of various performance indicators. By constructing a performance vector and combining the target performance requirements, the formula with the best comprehensive performance was selected as the final optimization solution. Subsequently, the optimized formula was also verified by field tests in a real swimming pool environment, further ensuring the stability and applicability of the formula in practical applications; Finally, on the basis of determining the optimal formula, the present invention also adjusts and optimizes the raw material ratio in a targeted manner. Using the performance change data under the actual application environment, combined with the aforementioned environmental parameter model, the formula is finely adjusted using a mathematical optimization algorithm, so that its performance indicators are more in line with the requirements for swimming pool use; In general, the optimization method proposed in the present invention fully explores the influence of various factors in the use environment of swimming pools on the performance of waterproof materials, and adopts scientific modeling, prediction, verification and other means to systematically optimize the formula design of polyurea waterproof materials in swimming pools. Compared with the prior art, the optimized formula obtained in the present invention has significantly improved in waterproofness, weather resistance, transparency, mechanical properties, construction performance and special performance, providing a better waterproof solution for swimming pools. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

[0034] Figure 1 The present invention is a flow chart of the optimization method for the two-component solvent-free brush-applied transparent waterproof polyurea formulation. DETAILED DESCRIPTION

[0035] In order to make the purpose, technical solution and advantages of the embodiments of the present invention more clear, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0036] like Figure 1 As shown, it is a flow chart of the first embodiment of the two-component solvent-free brush-coated transparent waterproof polyurea formula optimization method provided by the present invention. The specific implementation methods of the above steps are described in detail below: The specific implementation of step S10 is as follows: First, collect the operating environment data of multiple swimming pools. The water temperature change data can be modeled using a sine function model, that is, ,in for The water temperature at the moment, is the average water temperature, is the temperature fluctuation amplitude, is a random fluctuation term. The water pressure data can be calculated according to the water depth using the hydrostatic pressure relationship, that is, ,in For Depth The water pressure at is the water density, is the acceleration due to gravity, is the atmospheric pressure. In addition, water quality parameter data needs to be collected, such as residual chlorine content, pH value, turbidity, and microbial content in water.

[0037] Secondly, obtain the multimodal fingerprint data of each raw material in the two-component solvent-free brush-coated transparent waterproof polyurea formula for swimming pools. The multimodal fingerprint includes spectral data, morphological data, thermal analysis data and other characterization data. Spectral data can be obtained by infrared spectroscopy, Raman spectroscopy and other analysis to reflect the existence and strength of chemical bonds; morphological data can be obtained by scanning electron microscopy, atomic force microscopy and other means to reflect the microscopic morphology and scale distribution of raw materials; thermal analysis data can be obtained by thermogravimetric analysis, differential scanning calorimetry and other means to reflect the thermal stability and phase change behavior of raw materials; other characterization data such as X-ray diffraction, particle size analysis, etc., are used to reflect the structure, composition and particle size information of raw materials.

[0038] The similarity calculation of multimodal fingerprint data adopts the weighted average method, that is, ,in is the overall similarity index, For the The similarity index of the modalities, For the The specific similarity calculation formula is as follows: Spectral similarity ; Morphological similarity ; Thermal analysis similarity .

[0039] By collecting comprehensive swimming pool operating environment data and obtaining multimodal fingerprint data of various raw materials, the necessary basic information is provided for subsequent formula optimization.

[0040] The specific implementation of step S20 is as follows: According to the swimming pool use environment parameter model, materials with multimodal fingerprint data similarity greater than the preset threshold and meeting the environmental parameter requirements are selected as replacement raw materials. The specific environmental parameter matching degree calculation formula is: ,in For the The matching degree of environmental parameters of alternative raw materials, For the The matching function of the environment parameters, For the Alternative raw materials Environmental parameter values, is the total number of environmental parameters.

[0041] The screening conditions for alternative raw materials are ,in For the The multimodal fingerprint similarity of the alternative raw materials, is the preset fingerprint similarity threshold, is the environmental parameter matching degree, Matches the threshold value for the preset environmental parameters.

[0042] Under the premise of meeting the above conditions, the traversal loop method is used to generate the formula set, which can ensure that the formula set contains available raw material alternatives and also meets the parameter requirements of the swimming pool use environment.

[0043] The specific implementation of step S30 is as follows: First, a quantitative structure-activity relationship chemical model of swimming pool waterproof materials was established, namely ,in is the target performance indicator, is the ratio of each raw material, are model parameters, is the amount of raw materials. This allows a mathematical relationship between the formulation composition and the final performance to be established.

[0044] Then, combined with the swimming pool environment parameter model, the optimal ratio of raw materials is calculated for each formula in the generated formula set. The specific calculation conditions include: , that is, the total ratio of raw materials is 1; , that is, the ratio of each raw material needs to meet the upper and lower limit constraints. The raw material ratio obtained by optimization is the optimal ratio of the formula.

[0045] In this way, we can make full use of the existing environmental parameter models and structure-activity relationship models to calculate the optimal raw material ratios for each alternative formulation, providing basic data for subsequent performance prediction and screening.

[0046] The specific implementation of step S40 is as follows: First, for each recipe in the input recipe set, the features of its multimodal fingerprint data are extracted. Specifically, a convolutional neural network (CNN) is used to extract the features of spectral data and morphological data, namely and ; The long short-term memory network (LSTM) is used to extract the features of thermal analysis data, namely ; For other representation data, fully connected neural network is used to extract features.

[0047] Then, the attention mechanism is used to fuse the features of these different modalities to obtain the fused feature vector This enables interactive learning and feature fusion between multimodal data.

[0048] Finally, the fused feature vector and the raw material ratio vector Input into the multi-layer perceptron (MLP) to predict the swimming pool waterproof performance index of the formula .

[0049] Such a multi-level performance prediction network structure can make full use of the supplementary information of different modal data and comprehensively consider the impact of raw material ratio on performance, so as to obtain more accurate and reliable performance prediction results.

[0050] The specific implementation of step S50 is as follows: First, multiple formulations with the best swimming pool waterproof performance index are selected to form a first formulation set. For each formulation in the first formulation set, a small test is conducted in a laboratory simulating a swimming pool environment to obtain multiple necessary performance indicators such as the waterproof performance, mechanical properties, weather resistance, construction performance, and swimming pool-specific performance of the polyurea material.

[0051] Then, the above multiple performance indicators are standardized and constructed into a performance vector ,in to They represent the standardized values ​​of waterproof performance, mechanical properties, weather resistance, construction performance and special performance respectively. In this way, different performance characteristics can be mapped to the same dimension for comparative analysis.

[0052] Through the above steps, corresponding performance vectors are obtained for each recipe in the first recipe set, which provide important basis for subsequent recipe screening and optimization.

[0053] The specific implementation of step S60 is as follows: In a laboratory simulated swimming pool environment, each formula in the first formula set is subjected to an accelerated aging test to test its weather resistance. Specifically, an exponential decay model can be used to describe the change of performance over time, where for Performance value at the moment, is the initial performance value, is the aging rate constant.

[0054] The accelerated aging test can quickly obtain the weathering performance change data of each formula in a simulated swimming pool environment. Combined with other performance data obtained from the aforementioned small-scale experiment, the complete performance vector of each formula in the first formula set can be updated. This provides a more comprehensive performance evaluation basis for subsequent formula screening.

[0055] The specific implementation of step S70 is as follows: According to the performance vector of each recipe , calculate its performance vector with the preset target The Euclidean distance between ,in is the weight coefficient of each performance index.

[0056] Select distance The smallest formula is taken as the optimized formula. This ensures that the performance indicators of the optimized formula are closest to the target requirements and is the best comprehensive performance formula.

[0057] The specific implementation of step S80 is as follows: The optimized formula selected in step S70 is used to conduct field tests in a real swimming pool environment. During the test, it is necessary to comprehensively record the data of changes in material properties over time, including the change trends of various indicators such as waterproof performance, mechanical properties, and weather resistance.

[0058] These field verification data can provide a performance reference for subsequent formulation optimization in actual application environments. By comparing laboratory simulation results and field test data, the applicability and stability of the optimized formulation in actual applications can be further evaluated.

[0059] The specific implementation of step S90 is as follows: According to the material performance change data obtained in step S80, combined with the swimming pool use environment parameter model established above, the optimization formula is adjusted and optimized. The specific optimization method is: ,in For the optimized ratio, For the original ratio, is the adjustment factor, is the performance change rate.

[0060] In this way, the raw material ratio can be appropriately adjusted according to the performance in the actual application environment, so that the optimized formula can better adapt to the use environment of the swimming pool. Through iterative optimization, the comprehensive performance level and applicability of the optimized formula can be further improved.

[0061] In summary, the present invention adopts a series of scientific methods and algorithms to optimize the formulation of a two-component solvent-free brush-coated transparent waterproof polyurea for swimming pools, including: 1. Collect swimming pool operating environment data and obtain raw material multimodal fingerprint data to provide basic data support for formula optimization; 2. Screen suitable raw material alternatives based on environmental parameter matching and fingerprint similarity to generate a set of candidate formulas; 3. Use the structure-activity relationship model to calculate the optimal raw material ratio of each alternative formula; 4. Use a multi-level performance prediction network to comprehensively evaluate the performance of alternative formulations; 5. Screen the optimal formula based on the performance vector and conduct field tests to verify it in actual applications; 6. Further adjust and optimize the optimized formula based on field test data.

[0062] Among them, in the above technical scheme, the two-component solvent-free brush-applied transparent waterproof polyurea formula for swimming pools includes component A and component B, component A contains aliphatic isocyanate prepolymer and additives, wherein the additives include anti-ultraviolet agent, anti-settling agent, defoaming agent, and curing accelerator, component B contains polyether amine, amine chain extender and additives, wherein the additives include antioxidant, wetting dispersant, leveling agent, and toughening agent, and the mass ratio of component A to component B is 1:1 to 1:1.2.

[0063] Furthermore, in the above technical scheme, component A comprises 80 to 95 parts of aliphatic isocyanate prepolymer and 5 to 20 parts of additives by weight, wherein the additives comprise 1 to 3 parts of anti-ultraviolet agent, 0.5 to 2 parts of anti-settling agent, 0.3 to 1 part of defoaming agent, and 0.5 to 2 parts of curing accelerator; component B comprises 60 to 80 parts of polyetheramine, 15 to 25 parts of amine chain extender and 5 to 15 parts of additives by weight, wherein the additives comprise 1 to 2 parts of antioxidant, 0.5 to 1.5 parts of wetting and dispersing agent, 0.3 to 1 part of leveling agent, and 2 to 5 parts of toughening agent.

[0064] Further, in the above technical solution, multimodal fingerprint data refers to characteristic data of raw materials obtained by multiple analysis methods, including spectral data, morphological data, thermal analysis data and other characterization data; The spectral data includes one or more of infrared spectrum, Raman spectrum, ultraviolet spectrum and visible spectrum, which are used to reflect the existence and strength of chemical bonds; The morphology data includes one or more of scanning electron microscope images, atomic force microscope images, transmission electron microscope images, and optical microscope images, which are used to reflect the microscopic morphology and size distribution of the raw materials; Thermal analysis data include one or more of thermogravimetric analysis data, differential scanning calorimetry analysis data, thermomechanical analysis data, and dynamic mechanical thermal analysis data, which are used to reflect the thermal stability and phase change behavior of the raw materials; Other characterization data include one or more of X-ray diffraction data, particle size analysis data, specific surface area data, and density data, which are used to reflect the structure, composition, and particle size information of the raw materials.

[0065] Further, in the above technical solution, the performance vector is a vector composed of waterproof performance, mechanical performance, weather resistance performance, construction performance, and swimming pool-specific performance; Waterproof performance includes one or more of the following: anti-seepage strength, water absorption rate, water permeability, and waterproof sealing performance; Mechanical properties include one or more of tensile strength, elongation at break, tear strength, and underwater adhesion; Weather resistance includes one or more of resistance to chloride ion corrosion, resistance to ozone aging, resistance to hydrolysis, and resistance to temperature cycling; Construction performance includes one or more of surface drying time, curing time, coating thickness, and underwater construction adaptability; The special properties of swimming pools include one or more of resistance to disinfectants, anti-algae, anti-slip and transparency.

[0066] Furthermore, in the above technical scheme, the structure of the swimming pool waterproofing performance prediction model is a multi-level performance prediction network structure, including five parallel performance prediction modules. Each module realizes information sharing and feature fusion through an interactive learning mechanism. The performance prediction modules include a basic waterproofing performance prediction module, a mechanical performance prediction module, an environmental adaptability prediction module, a construction performance prediction module and a special performance prediction module.

[0067] Further, in the above technical solution, the basic waterproof performance prediction module mainly predicts the basic waterproof performance index of the material; The mechanical properties prediction module predicts the mechanical properties of materials based on the basic waterproof performance characteristics; The environmental adaptability prediction module integrates the waterproof performance and mechanical performance characteristics to predict the weathering performance index of the material; The construction performance prediction module comprehensively considers the aforementioned performance characteristics to predict the construction application performance index of the material; The dedicated performance prediction module integrates all performance characteristics to predict the dedicated performance indicators of the material in the specific environment of the swimming pool.

[0068] Furthermore, in the above technical solution, each prediction module includes five network modules: a spectral data prediction network module, a morphological data prediction network module, a thermal analysis data prediction network module, other characterization data prediction network module and a multimodal fusion network module.

[0069] Further, in the above technical solution, the spectral data prediction network module is used to extract the characteristics of the spectral data of the spectral data using a convolutional neural network or a recurrent neural network structure, and predict the performance index corresponding to the characteristics of the spectral data; The morphology data prediction network module is used to extract the morphology features of the morphology data using a convolutional neural network and predict the performance indicators corresponding to the morphology features; The thermal analysis data prediction network module is used to extract the thermal analysis features of the thermal analysis data using a fully connected neural network or a long short-term memory network structure, and predict the performance indicators corresponding to the thermal analysis features; The other representation data prediction network module is used to extract other representation features of other representation data using a fully connected neural network or a long short-term memory network structure, and predict the performance indicators corresponding to other representation features; The multimodal fusion network module is used to fuse different spectral data prediction network modules, morphological data prediction network modules, thermal analysis data prediction network modules, and other characterization data prediction network modules using an attention mechanism or a gating mechanism, and ultimately predict all performance indicators corresponding to all multimodal fingerprint data.

[0070] Furthermore, in the above technical solution, the water quality parameter data includes one or more of residual chlorine content, pH value, turbidity, and microbial content in water.

[0071] The second embodiment provided by the present invention: A swimming pool in a certain city is planning to renovate and upgrade the original swimming pool, and requires the use of a new type of waterproof material. The swimming pool is located in the central area of ​​the city and is an open-air swimming pool. The peak usage period is from May to October each year. According to past operating data statistics, the water temperature of the swimming pool ranges from 20 to 30°C, the water depth is between 1.5 and 2.5 meters, the water quality is good, the pH value is between 7.2 and 7.8, the residual chlorine content is 0.52 mg / L, the turbidity is less than 5 NTU, and no harmful microbial contamination is detected. From June to September each year, the ultraviolet radiation intensity in the area where the swimming pool is located is relatively high, and the average sunshine intensity is 35 to 45 W / m².

[0072] In order to meet the waterproof performance requirements of the swimming pool, the R&D team decided to use the optimization method proposed in this invention to design a two-component solvent-free brush-coated transparent waterproof polyurea formula. First, through on-site investigation and historical data analysis, the use environment parameter model of the swimming pool was established. The details are as follows: The water temperature change model is: ; In the formula, is the water temperature at time t, in °C; is a random fluctuation term, which follows a normal distribution .

[0073] The water pressure model is: ; In the formula, For Depth The water pressure at the point, in Pa; Take 1000kg / m³; Take 9.81m / s².

[0074] The statistics of water quality parameters are shown in Table 1: Table 1 Statistics of swimming pool water quality parameters parameter Value range pH 7.2~7.8 Residual chlorine content 0.5~2.0mg / L Turbidity <5NTU Microbial Content No contamination detected The UV radiation intensity model is: ; In the formula, is the sunshine intensity at time t, in W / m²; is a random fluctuation term, which follows a normal distribution .

[0075] Next, the R&D team obtained multimodal fingerprint data of several commonly used polyurea raw materials on the market, including spectral data, morphological data, thermal analysis data and other characterization data. Taking raw material A as an example, its characteristic data is shown in Table 2: Table 2A Raw material multimodal fingerprint data Feature Type Feature data Spectral data The infrared spectrum absorption peaks are located at 2270cm{-1}, 1730cm{-1}, 1530cm^{-1}, etc. Morphological data Scanning electron microscope observation showed irregular particles with a particle size distribution of 10~50μm Thermal analysis data Thermogravimetric analysis showed that the weight loss began to be obvious above 300°C, and differential scanning calorimetry showed that the glass transition temperature was about -20°C. Other manifestations The X-ray diffraction peaks are located at 2θ=11.3°, 19.7°, 26.9°, etc., and the specific surface area is about 25m² / g Then, the R&D team calculated the matching degree of the environmental parameters of raw material A based on the environmental parameters of the swimming pool. According to the above formula: ; in, They are the matching functions of environmental parameters such as temperature, pressure, water quality, and ultraviolet light. .

[0076] Similarly, similar environmental parameter matching analysis was performed on several other polyurea raw materials on the market. Combined with the multimodal fingerprint similarity calculation of each raw material, a list of alternative raw materials that meet the environmental parameter requirements and have a fingerprint similarity greater than 0.8 was screened out, including raw material B, raw material C, and raw material D.

[0077] Based on the above screening results, the R&D team began to optimize the formula design. First, a quantitative structure-activity relationship model was established to mathematically describe the relationship between the ratio of various raw materials in the formula and the final performance. After multiple rounds of experimental verification, the following model was finally obtained: ; In the formula, represents the target performance indicator value; They are the ratios of the four raw materials A, B, C, and D. At the same time, the upper and lower limits of the ratio of each raw material are set, that is, and .

[0078] On this basis, the R&D team used a multi-level performance prediction network to conduct a comprehensive evaluation of the alternative formulas. First, the spectrum, morphology, thermal analysis and other features of each raw material were extracted, and the features were extracted using structures such as convolutional neural networks (CNN) and long short-term memory networks (LSTM). Then, the attention mechanism was used to fuse the features of different modes to obtain a comprehensive feature vector. Finally, the fused features were input into the multi-layer perceptron (MLP) network together with the raw material ratio to predict various performance indicators, including waterproofness, mechanical properties, weather resistance, construction properties and swimming pool-specific performance. Through the collaborative learning of different network modules, the multi-level prediction network finally achieved accurate prediction of formula performance.

[0079] After the above modeling and prediction, the R&D team initially screened out several alternative formulations with better comprehensive performance. Their performance indicators are shown in Table 3: Table 3 Prediction results of performance indicators of alternative formulations

[0080] Among them, the performance indicators are all standardized 5-point scores. As can be seen from the table, the comprehensive performance indicators of formula 3 are relatively optimal and meet all the requirements of swimming pool waterproof materials.

[0081] To further verify the actual performance of these alternative formulas, the R&D team conducted a small test in a laboratory simulating a swimming pool environment. Specifically: 1. Test the basic waterproof properties of each formula, such as anti-seepage strength, water absorption rate, and water permeability, in a water temperature range of 20~30℃; 2. Test mechanical properties such as tensile strength, elongation at break, underwater adhesion, etc.; 3. Carry out weather resistance tests such as resistance to chloride ion corrosion and resistance to ozone aging; 4. Test construction performance indicators such as surface drying time, curing time, and underwater construction adaptability; 5. Test the special performance of swimming pools such as resistance to disinfectants, algae resistance, transparency, etc.

[0082] Through the above comprehensive performance tests, the R&D team obtained multiple performance data of each candidate formula in the actual simulation environment, and constructed a performance vector according to the pre-set target performance requirements. After calculation, the performance vector of formula 3 is ; With target performance vector The Euclidean distance is the smallest, so formula 3 is determined to be the optimal formula.

[0083] Next, the R&D team conducted field tests to verify Formula 3 in a real swimming pool environment. During the six-month test period, the performance indicators of the material were fully monitored. The results showed that all performance indicators met the requirements, and no leakage, bulging, peeling, etc. were found. At the same time, the material's special properties such as transparency and anti-slip properties were also well evaluated.

[0084] Based on the results of laboratory simulation and field tests, the R&D team believes that Formula 3 can meet the waterproof material requirements of the swimming pool. However, in order to further optimize the performance, the formula was fine-tuned in a targeted manner. Specifically, based on the field test data, the change rate of each performance indicator was calculated, and combined with the aforementioned environmental parameter model, the following formula was used to adjust the raw material ratio: ; In the formula, After optimization Ratio of raw materials; For the original ratio; For the The change rate of the corresponding performance indicators of the raw materials.

[0085] After two rounds of iterative optimization, the optimized formula ratio was finally obtained, as shown in Table 4: Table 4 Optimized formula ratio raw material Initial ratio Optimized ratio A Raw materials 0.35 0.33 B Raw Materials 0.25 0.27 C Raw Materials 0.25 0.24 D Raw materials 0.15 0.16 The performance indicators of the optimized formula in terms of waterproofness, mechanical properties, weather resistance, construction performance and special purpose all reach or exceed 4.9 points out of 5 points, fully meeting the use requirements of the swimming pool.

[0086] Through the above optimization design process, the swimming pool finally adopts the two-component solvent-free brush-coated transparent waterproof polyurea formula proposed by the present invention. This formula performs well in key properties such as waterproofness, weather resistance, and transparency, and also has good mechanical properties, construction performance and swimming pool-specific performance, providing a reliable waterproof solution for swimming pools. Compared with existing PVC waterproof membranes, epoxy resin coatings, etc., this polyurea waterproof material is not only more environmentally friendly, but also simple to construct, which can effectively extend the service life of the swimming pool and create a safe and comfortable swimming environment for swimmers.

[0087] Table 5 Explanation of variables in the invention

[0088] Specifically, the principle of the present invention is: the core of the two-component solvent-free brush-coated transparent waterproof polyurea formula optimization method proposed in the present invention is to make full use of data-driven modeling and prediction technology to systematically optimize the formula design: First, by collecting the actual operating environment data of the swimming pool, a mathematical model describing the changing laws of parameters such as water temperature, water pressure, and water quality was established. This provides the necessary environmental condition constraints for subsequent raw material screening and formula optimization. At the same time, multimodal fingerprint data of various raw materials are obtained, including spectrum, morphology, thermal analysis, etc., and corresponding feature extraction and similarity calculation methods are established. Through the dual screening of environmental parameter matching and fingerprint similarity, the most suitable alternative raw materials for swimming pools can be selected from a large number of raw materials, laying the foundation for formula design; Secondly, in view of the performance characteristics of swimming pool waterproof materials, the present invention has established a quantitative structure-activity relationship model to quantify the mathematical relationship between the formula composition and the final performance. Combined with the aforementioned environmental parameter model, the optimal raw material ratio of each alternative formula is obtained through mathematical optimization. At the same time, a multi-level performance prediction network is used to comprehensively analyze the impact of multiple modal characteristics on performance, and predict the comprehensive performance indicators of each formula in the swimming pool environment, including waterproofness, mechanical properties, weather resistance, construction and specificity. This provides a reliable basis for subsequent formula screening and optimization; Thirdly, the present invention conducted a small test and accelerated aging test on the candidate formula in a laboratory simulated swimming pool environment, and obtained the measured data of various performance indicators. By constructing a performance vector and comparing it with the target performance requirements, the formula with the best comprehensive performance can be screened out. At the same time, field tests were conducted in a real swimming pool environment to ensure the stability and applicability of the optimized formula in practical applications; Finally, on the basis of determining the optimal formula, the present invention also adjusts and optimizes the raw material ratio in a targeted manner. Using the performance change data under the actual application environment, combined with the aforementioned environmental parameter model, the formula is finely adjusted using a mathematical optimization algorithm, so that its performance indicators are more in line with the requirements for swimming pool use. This can further improve the comprehensive performance level and applicability of the optimized formula; In general, the optimization method proposed in the present invention mainly includes the following key technical innovations: 1. Comprehensively consider the environmental parameters of swimming pool use, establish an accurate environmental parameter model, and provide a basis for raw material screening and formula optimization; 2. Use multimodal feature fingerprint data to select the most suitable raw materials for swimming pools through similarity matching; 3. Establish a quantitative structure-activity relationship model and a multi-level performance prediction network to accurately predict formulation performance and provide support for optimization screening; 4. Combine laboratory simulation and field tests to systematically verify the comprehensive performance and stability of the optimized formula; 5. Based on the actual application environment data, mathematical optimization algorithms are used to fine-tune the formula to further improve the performance level.

Claims

1. A two-component solvent-free brush-coated transparent waterproof polyurea formulation optimization method, characterized in that: The following steps are involved: S10, collecting operating environment data of multiple swimming pools, including water temperature change data, water pressure data, water quality parameter data, and ultraviolet irradiation intensity data; obtaining multimodal fingerprint data of each raw material in a two-component solvent-free brush-coated transparent waterproof polyurea formula for swimming pools; S20, establishing a swimming pool use environment parameter model according to the swimming pool operation environment data, for each raw material, selecting a material whose multimodal fingerprint data similarity with the raw material is greater than a preset fingerprint similarity threshold and meets the requirements of the swimming pool use environment parameter model as a substitute raw material, and generating a recipe set by using a traversal loop method; S30, using a quantitative structure-activity relationship chemical model for swimming pool waterproof materials and combining it with the swimming pool use environment parameter model, calculating the raw material ratio of each formula in the formula set; S40, using a pre-trained swimming pool-specific waterproof performance prediction model, inputting a multimodal fingerprint and a raw material ratio of each formula in the formula set, and obtaining a swimming pool waterproof performance index of each formula; S50, selecting multiple formulas with the best swimming pool waterproof performance index to form a first formula set, and obtaining multiple necessary properties of the polyurea material obtained by conducting a small test on each formula in the first formula set, and generating a performance vector; S60, performing an accelerated aging test in a laboratory simulating a swimming pool environment to test the aging resistance of each formulation in the first formulation set, and updating the performance vector; S70, according to the performance vector of each formula, selecting the formula closest to the required performance vector of the preset two-component solvent-free brush-coated transparent waterproof polyurea for swimming pools as the optimized formula; S80, using the optimized formula to conduct field tests in a real swimming pool environment, and recording material performance change data; S90. According to the material performance change data and in combination with the swimming pool use environment parameter model, the raw material ratio of the optimized formula is adjusted to obtain a final formula.

2. The two-component solvent-free brush-coated transparent waterproof polyurea formulation optimization method according to claim 1, characterized in that: The two-component solvent-free brush-coated transparent waterproof polyurea formula for swimming pools comprises component A and component B, wherein component A comprises aliphatic isocyanate prepolymer and additives, wherein the additives comprise an anti-ultraviolet agent, an anti-settling agent, a defoaming agent, and a curing accelerator, and component B comprises polyetheramine, an amine chain extender and additives, wherein the additives comprise an antioxidant, a wetting and dispersing agent, a leveling agent, and a toughening agent, and the mass ratio of component A to component B is 1:1 to 1:1.

2.

3. The two-component solvent-free brush-coated transparent waterproof polyurea formulation optimization method according to claim 2, characterized in that: The component A comprises 80 to 95 parts of the aliphatic isocyanate prepolymer and 5 to 20 parts of the auxiliary agent by weight, wherein the auxiliary agent comprises 1 to 3 parts of the anti-ultraviolet agent, 0.5 to 2 parts of the anti-settling agent, 0.3 to 1 part of the defoaming agent, and 0.5 to 2 parts of the curing accelerator; the component B comprises 60 to 80 parts of the polyether amine, 15 to 25 parts of the amine chain extender, and 5 to 15 parts of the auxiliary agent by weight, wherein the auxiliary agent comprises 1 to 2 parts of the antioxidant, 0.5 to 1.5 parts of the wetting and dispersing agent, 0.3 to 1 part of the leveling agent, and 2 to 5 parts of the toughening agent.

4. The two-component solvent-free brush-coated transparent waterproof polyurea formulation optimization method according to claim 3, characterized in that: The multimodal fingerprint data refers to characteristic data of raw materials obtained through various analytical methods, including spectral data, morphological data, thermal analysis data and other characterization data; The spectral data includes one or more of infrared spectrum, Raman spectrum, ultraviolet spectrum and visible spectrum, which are used to reflect the existence and strength of chemical bonds; The morphology data includes one or more of a scanning electron microscope image, an atomic force microscope image, a transmission electron microscope image, and an optical microscope image, which are used to reflect the microscopic morphology and size distribution of the raw material; The thermal analysis data includes one or more of thermogravimetric analysis data, differential scanning calorimetry analysis data, thermomechanical analysis data, and dynamic mechanical thermal analysis data, which are used to reflect the thermal stability and phase change behavior of the raw materials; The other characterization data include one or more of X-ray diffraction data, particle size analysis data, specific surface area data, and density data, which are used to reflect the structure, composition and particle size information of the raw material.

5. The two-component solvent-free brush-coated transparent waterproof polyurea formulation optimization method according to claim 4, characterized in that: The performance vector is a vector composed of waterproof performance, mechanical performance, weather resistance performance, construction performance, and swimming pool-specific performance; The waterproof performance includes one or more of the following: anti-seepage strength, water absorption rate, water permeability, and waterproof sealing performance; The mechanical properties include one or more of tensile strength, elongation at break, tear strength, and underwater adhesion; The weather resistance includes one or more of chloride ion corrosion resistance, ozone aging resistance, hydrolysis resistance, and temperature cycle resistance; The construction performance includes one or more of surface drying time, curing time, coating thickness, and underwater construction adaptability; The swimming pool-specific properties include one or more of resistance to disinfectants, algae resistance, slip resistance, and transparency.

6. The two-component solvent-free brush-coated transparent waterproof polyurea formulation optimization method according to claim 5, characterized in that: The structure of the swimming pool waterproof performance prediction model is a multi-level performance prediction network structure, including five parallel performance prediction modules. Each module realizes information sharing and feature fusion through an interactive learning mechanism. The performance prediction modules include a basic waterproof performance prediction module, a mechanical performance prediction module, an environmental adaptability prediction module, a construction performance prediction module and a special performance prediction module.

7. The method for optimizing the formulation of the two-component solvent-free brush-coated transparent waterproof polyurea according to claim 6, characterized in that: The basic waterproof performance prediction module mainly predicts the basic waterproof performance index of the material; The mechanical property prediction module predicts the mechanical property index of the material in combination with the basic waterproof performance characteristics; The environmental adaptability prediction module integrates waterproof performance and mechanical performance characteristics to predict the weather resistance performance index of the material; The construction performance prediction module comprehensively considers the aforementioned performance characteristics to predict the construction application performance index of the material; The dedicated performance prediction module integrates all performance characteristics to predict the dedicated performance index of the material in the specific environment of the swimming pool.

8. The method for optimizing the formulation of the two-component solvent-free brush-coated transparent waterproof polyurea according to claim 7, characterized in that: Each prediction module includes five network modules: spectral data prediction network module, morphological data prediction network module, thermal analysis data prediction network module, other characterization data prediction network module and multimodal fusion network module.

9. The two-component solvent-free brush-coated transparent waterproof polyurea formulation optimization method according to claim 8, characterized in that: The spectral data prediction network module is used to extract the characteristics of the spectral data using a convolutional neural network or a recurrent neural network structure, and predict the performance index corresponding to the characteristics of the spectral data; The morphology data prediction network module is used to extract the morphology features of the morphology data using a convolutional neural network, and predict the performance indicators corresponding to the morphology features; The thermal analysis data prediction network module is used to extract thermal analysis features of thermal analysis data using a fully connected neural network or a long short-term memory network structure, and predict performance indicators corresponding to the thermal analysis features; The other characterization data prediction network module is used to extract other characterization features of other characterization data using a fully connected neural network or a long short-term memory network structure, and predict performance indicators corresponding to the other characterization features; The multimodal fusion network module is used to adopt an attention mechanism or a gating mechanism to fuse the different spectral data prediction network modules, morphological data prediction network modules, thermal analysis data prediction network modules, and other characterization data prediction network modules, and finally predict all performance indicators corresponding to all multimodal fingerprint data.

10. The method for optimizing the formulation of a two-component solvent-free brush-coated transparent waterproof polyurea according to claim 9, characterized in that: The water quality parameter data include one or more of residual chlorine content, pH value, turbidity, and microbial content in water.

Images