Integrated manufacturing method for optimizing the properties of water-based wear-resistant substitute finishing varnish

Abstract

The application provides a comprehensive manufacturing method for performance optimization of water-based wear-resistant substitute plastic finishing oil, comprising: obtaining wear resistance performance data and flexibility data of a paint sample, removing noise through preprocessing to obtain a standardized data set; according to the standardized data set, using a support vector machine to classify the relationship between wear resistance performance and flexibility, and determining boundary conditions of performance conflicts; extracting a feature vector from the boundary conditions of performance conflicts, and obtaining a high conflict point set through Euclidean distance judgment of high conflict points; for the high conflict point set, using a genetic algorithm to generate and evolve a candidate formula design scheme; evaluating the wear resistance performance and flexibility of the candidate formula design scheme through simulation testing to obtain a comprehensive score vector; obtaining a material stability index from the preliminary optimization result, and determining a final formula design scheme through comparison with original paint data.

Description

Technical Field

[0001] This invention relates to the field of information technology, and in particular to a comprehensive manufacturing method for optimizing the performance of water-based abrasion-resistant plastic substitute varnishes. Background Technology

[0002] In modern manufacturing, the research and application of environmentally friendly coating materials has become an important direction for promoting green development and industrial upgrading. Especially in the packaging, printing, and consumer electronics sectors, water-based varnishes, due to their low pollution and low toxicity, are considered ideal alternatives to traditional oil-based coatings and possess undeniable strategic value. However, despite the significant attention this field has received, its technological development and practical application face numerous challenges, urgently requiring breakthrough innovations to meet the growing market demand.

[0003] Current waterborne varnish technology has significant limitations in performance improvement. Many existing solutions often fail to find a balance among multiple key performance indicators, especially when facing complex application scenarios, exhibiting obvious shortcomings. The problem lies not only in the performance defects of the materials themselves, but also in the lack of a systematic design approach, making it difficult for products to meet diverse needs in actual use. This limitation restricts the competitiveness of waterborne varnishes in the high-end market, especially in scenarios requiring a balance of multiple performance aspects.

[0004] Focusing further on the technical challenges, the core obstacle hindering the development of water-based varnishes lies in the contradiction between abrasion resistance and flexibility. Abrasion resistance, a crucial indicator of a coating material's long-term durability, directly impacts product longevity, while flexibility determines whether the coating will crack or fail when bent or deformed. This inherent conflict means that many products, while pursuing high abrasion resistance, often lose necessary elasticity due to overly rigid coatings, making them unable to withstand repeated physical stresses in practical applications. For example, in flexible packaging, coatings need to maintain surface integrity and protective function while withstanding multiple folds, but existing technologies often struggle to meet both requirements simultaneously. Coatings either crack due to excessive hardness or suffer severe wear due to insufficient flexibility.

[0005] Therefore, achieving an effective balance between abrasion resistance and flexibility in the formulation design of water-based varnishes, while ensuring the coating remains stable over a long period in complex operating environments, has become a key issue that this research urgently needs to address. Solving this problem not only concerns the improvement of material performance but also directly impacts whether water-based varnishes can truly replace traditional plastic films and meet the demands of green manufacturing and high-end applications. Summary of the Invention

[0006] This invention provides a comprehensive manufacturing method for optimizing the performance of water-based abrasion-resistant plastic substitute varnishes, mainly including:

[0007] Abrasion resistance and flexibility data of coating samples are obtained, and noise is removed through preprocessing to obtain a standardized dataset. Based on the standardized dataset, a support vector machine is used to classify the relationship between abrasion resistance and flexibility, and the boundary conditions of performance conflicts are determined. Feature vectors are extracted from the boundary conditions of performance conflicts, and high conflict points are obtained by identifying high conflict points using Euclidean distance. For the high conflict point set, a genetic algorithm is used to generate and evolve alternative formulation design schemes. The abrasion resistance and flexibility of the alternative formulation design schemes are evaluated through simulation tests to obtain a comprehensive score vector. The relationship between the average value of the comprehensive score vector and a preset threshold is used to determine whether to iterate the genetic algorithm; otherwise, the highest-scoring scheme is selected as the preliminary optimization result. Material stability indicators are obtained from the preliminary optimization results, and the final formulation design scheme is determined by comparing them with the original coating data. Furthermore, the acquisition of abrasion resistance and flexibility data of coating samples, and the obtaining of a standardized dataset through preprocessing to remove noise, includes: constructing an initial dataset by acquiring raw data of abrasion resistance and flexibility values ​​from coating samples; if the noise in the initial dataset exceeds a preset threshold, preprocessing is performed using a filtering method to obtain a denoised dataset; standardizing the abrasion resistance and flexibility values ​​according to the denoised dataset to generate a standardized dataset; if the performance indicators in the standardized dataset meet preset conditions, recording quality assessment parameters to obtain assessment criteria; classifying the abrasion resistance and flexibility values ​​in the standardized dataset using a support vector machine based on the assessment criteria to determine the classification results; constructing an analysis model based on the classification results to obtain performance prediction output data; and adjusting the analysis model parameters to determine the performance distribution if there is a deviation between the performance prediction output data and the quality assessment parameters. Furthermore, the step of classifying the relationship between wear resistance and flexibility using support vector machines based on the standardized dataset to determine the boundary conditions of performance conflicts includes: obtaining records of wear resistance and flexibility from the standardized dataset, performing preliminary screening using data processing methods to obtain a set of organized performance indicators; classifying wear resistance and flexibility separately based on the organized performance indicator set, using support vector machines to divide the interaction between the two to determine preliminary classification results; analyzing the conflict boundaries in the performance indicators through the preliminary classification results to identify inconsistencies between wear resistance and flexibility and obtain the distribution range of conflict boundaries; processing data on potential conflict areas based on the distribution range of conflict boundaries; marking areas where performance indicators deviate from preset thresholds; obtaining key points through the marked conflict areas; performing in-depth comparisons on these points to determine the conflicting performance of wear resistance and flexibility under specific conditions; and storing the inconsistent parts in the classification results based on the conflicting performance using data recording tools to obtain a performance analysis archive.Furthermore, the step of extracting feature vectors from the boundary conditions of the performance conflict and obtaining a set of high conflict points by judging high conflict points through Euclidean distance includes: obtaining relevant data on performance conflicts from the boundary conditions, performing core extraction on the data to obtain a set of feature vectors, calculating the spatial difference value between each vector using the Euclidean distance method on the feature vector set, performing conflict judgment to mark high conflict points and constructing a preliminary list of high conflict points by performing marking processing on the preliminary list of high conflict points to obtain classified conflict point groups and judge their distribution characteristics, performing set construction based on the conflict point grouping results to obtain a set of high conflict points, and obtaining the correlation data between the high conflict point set and the optimization basis to determine the direction of subsequent processing. Furthermore, the step of generating and evolving alternative formulation design schemes using a genetic algorithm for the set of high conflict points includes: obtaining internal set data for the set of high conflict points and processing it through data classification to obtain structured conflict data groups; generating an initial population based on the structured conflict data groups using a genetic algorithm to determine a preliminary formulation design framework; processing individual data in the initial population through crossover operations to obtain a recombined design individual set based on the preliminary formulation design framework; performing mutation operations on the recombined design individual set to introduce a random adjustment mechanism to obtain diverse evolutionary individual combinations; obtaining intermediate state data during the evolution process from the diverse evolutionary individual combinations and determining an optimized alternative scheme list through iterative screening; obtaining matching degree data between the optimized alternative scheme list and the formulation design; if the matching degree is lower than a preset threshold, performing further scheme evolution processing to obtain the final design population result; and integrating the final design population result through a scheme generation module to obtain a set of formulation design schemes. Furthermore, the step of evaluating the wear resistance and flexibility of the alternative formulation design schemes through simulation testing to obtain a comprehensive score vector includes: acquiring raw test data on the wear resistance and flexibility of each scheme through a simulation testing platform; normalizing the wear resistance and flexibility data using a standardization method based on the raw test data to obtain performance values ​​under a unified dimension; calculating the weighted results of wear resistance and flexibility using a pre-established weight allocation model to determine the comprehensive performance score of each scheme based on the performance values; if the comprehensive performance score is lower than a preset threshold, performing a secondary analysis on the performance values ​​of the scheme to obtain its deviation data in wear resistance and flexibility; classifying the performance distribution of the alternative schemes based on the deviation data using a support vector machine algorithm to determine the classification of each scheme in terms of performance balance; constructing a score vector for each scheme based on the classification results and the comprehensive performance score to obtain quantified performance balance level data; and generating a ranking result for each alternative scheme based on the quantified performance balance level data to determine the priority sequence of the optimal formulation design scheme.Furthermore, the step of determining whether to iterate the genetic algorithm based on the relationship between the average value of the comprehensive score vector and a preset threshold, and otherwise selecting the highest-scoring scheme as the preliminary optimization result, includes: obtaining comprehensive score data from multiple design schemes to construct a score vector set; calculating the mean of the vectors based on the initial score vector set; standardizing the mean using statistical tools to determine whether the mean reaches a preset threshold; if the mean of the vectors is lower than the preset threshold, triggering the genetic algorithm to iteratively adjust the formula design to generate a new set of design schemes; obtaining comprehensive score data again for the new set of design schemes to construct an updated score vector set; if the mean of the updated score vectors is still lower than the preset threshold, continuing to adjust the schemes through the genetic algorithm until the mean reaches or exceeds the preset threshold; and when the mean of the score vectors reaches or exceeds the preset threshold, selecting the scheme with the highest score as the preliminary optimization result and outputting the design scheme. Furthermore, the step of obtaining material stability indicators from the preliminary optimization results and determining the final formulation design scheme by comparing them with the original coating data includes: obtaining material stability indicator data from the preliminary optimization results, classifying and organizing each indicator using data extraction technology to obtain a structured indicator dataset; comparing and analyzing the structured indicator dataset with the original coating data, identifying the difference values ​​between the two using data matching methods to determine the difference dataset; using the support vector machine algorithm to classify and predict the relevant indicators of wear resistance and flexibility based on the difference dataset to obtain a preliminary evaluation result of performance balance; if the wear resistance index is higher than the flexibility index in the preliminary evaluation result, adjusting the weights of the formulation parameters and recalculating the performance balance value to obtain adjusted evaluation data; generating multiple alternative formulation combinations based on the adjusted evaluation data and the constraints of the final formulation design scheme to determine an alternative formulation list; obtaining simulated test data for each formulation in the alternative formulation list to analyze its comprehensive performance in wear resistance and flexibility to determine the optimal formulation scheme; and generating final formulation design output data based on the optimal formulation scheme to complete the performance balance indicator evaluation process. Furthermore, the process of generating and evolving alternative formulation design schemes using a genetic algorithm and evaluating the wear resistance and flexibility of the alternative formulation design schemes through simulation testing to obtain a comprehensive score vector includes: generating an initial population using a genetic algorithm for a set of high conflict points, and then evolving it through crossover and mutation operations to obtain diverse alternative formulation design schemes; obtaining original test data for the alternative formulation design schemes through simulation testing and performing normalization processing to obtain performance values; calculating a weighted comprehensive performance score based on the performance values ​​and constructing a score vector; if the comprehensive performance score or the average value of the score vector is lower than a preset threshold, iterating the genetic algorithm to adjust the formulation design scheme until the threshold condition is met; determining the highest-scoring scheme through the score vector as the preliminary optimization result, and further comparing the material stability index with the original coating data to determine the final formulation design scheme.

[0008] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects:

[0009] This invention discloses a comprehensive manufacturing method for optimizing the performance of water-based abrasion-resistant plastic substitute varnishes, aiming to solve the problem of balancing material properties in business scenarios, namely, finding the optimal trade-off between abrasion resistance and flexibility to avoid performance conflicts. This invention acquires and preprocesses sample data to construct a standardized dataset, then uses support vector machine classification technology to identify performance conflict boundaries, extracting key feature vectors and marking high-conflict points. Based on this, a genetic algorithm is applied to generate and evolve formulation design schemes, and the degree of performance balance is evaluated through simulation testing. Iterative optimization is performed until a preset scoring threshold is met, and finally, the optimal formulation is determined by comparing material stability indicators. This invention, through the combination of data-driven approaches and intelligent algorithms, accurately locates conflict areas and provides scientific formulation adjustment schemes, significantly improving the overall performance of water-based varnishes and providing an efficient and replicable optimization path for the field of materials design. Attached Figure Description

[0010] Figure 1 This is a flowchart of the comprehensive manufacturing method for optimizing the performance of water-based abrasion-resistant plastic substitute varnish according to the present invention. Detailed Implementation

[0011] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0012] like Figure 1 The comprehensive manufacturing method for optimizing the performance of water-based abrasion-resistant plastic substitute varnish in this embodiment may specifically include:

[0013] S101. Obtain the abrasion resistance and flexibility data of the water-based varnish samples. Preprocess these data to remove noise and obtain a standardized dataset for subsequent analysis.

[0014] An initial dataset is constructed by obtaining raw data on abrasion resistance and flexibility values ​​from water-based varnish samples. If the noise in the initial dataset exceeds a preset threshold, a filtering method is used to preprocess the data, resulting in a denoised dataset. Based on the denoised dataset, abrasion resistance and flexibility values ​​are standardized separately to generate standardized datasets. If the performance indicators in the standardized dataset meet preset conditions, the quality assessment parameters are recorded through the data acquisition module to obtain the assessment criteria. Based on the assessment criteria, a support vector machine algorithm is used to classify the abrasion resistance and flexibility values ​​in the standardized dataset, determining the classification results. Based on the classification results, an analysis model is constructed to meet the needs of subsequent analysis, obtaining performance prediction output data. If there is a deviation between the performance prediction output data and the quality assessment parameters, the parameters of the analysis model are adjusted to determine the final performance distribution.

[0015] To obtain abrasion resistance and flexibility data for water-based varnish samples, and to preprocess the data to remove noise, ultimately obtaining a standardized dataset for subsequent analysis, the following specific method can be used. First, assume that abrasion resistance data (expressed as the number of friction cycles, ranging from 5000 to 15000) and flexibility data (expressed as the bending angle, ranging from 30 to 120 degrees) for 100 water-based varnish samples are obtained through laboratory testing equipment. The data is stored in a CSV file, containing three columns: sample number, number of abrasion cycles, and bending angle. Next, the data is read using the Pandas library in the Python programming language, and outliers are detected using box plots. The upper and lower limits are set to 1.5 times the interquartile range. Outliers with fewer than 3000 or more than 18000 abrasion cycles, and bending angles less than 10 degrees or greater than 150 degrees are removed. Assume that 90 samples remain after this process. Then, to address the data noise issue, a moving average filtering algorithm was used to smooth the wear resistance count and bending angle data with a window size of 5, reducing the impact of random fluctuations. For example, if the original wear resistance count for a sample was 8000, it might be adjusted to 8050 after smoothing. Next, data standardization was performed using the StandardScaler method from the Scikit-learn library to convert the wear resistance count and bending angle data into a standard normal distribution with a mean of 0 and a standard deviation of 1. For example, the standardized wear resistance count for a sample was 0.25, and the bending angle was -0.18, ensuring the comparability of data with different dimensions. Finally, the standardized dataset was saved as a new CSV file, and a scatter plot was drawn using visualization tools such as Matplotlib to analyze the correlation between wear resistance and flexibility. Assuming a Pearson correlation coefficient of 0.65, it indicates a positive correlation, providing data support for subsequent modeling. Through this series of processes, from raw data collection to standardized output, a complete data processing chain was formed, ensuring data quality and laying the foundation for further performance optimization analysis.

[0016] S102. Based on the standardized dataset, support vector machines are used to classify the relationship between wear resistance and flexibility, determine the boundary conditions of performance contradictions, and identify potential conflict areas.

[0017] Step 1: Obtain the original records of wear resistance and flexibility properties from the standardized dataset. Use data processing methods to initially screen the records, resulting in a set of organized performance indicators. Step 2: Based on the organized performance indicator set, classify wear resistance and flexibility properties separately. Use support vector machines to partition the interaction between the two, determining the preliminary classification results. Step 3: Analyze the conflict boundaries in the performance indicators based on the preliminary classification results, identifying inconsistencies between wear resistance and flexibility properties, and obtaining the distribution range of the conflict boundaries. Step 4: Based on the distribution range of the conflict boundaries, further process the data for potential conflict areas. If the performance indicators in a certain area deviate from a preset threshold, mark it to determine the marked conflict area. Step 5: Through the marked conflict areas, obtain key points related to performance analysis. Conduct in-depth comparisons of these points to determine the conflicting performance of wear resistance and flexibility properties under specific conditions. Step 6: Based on the in-depth comparison of conflicting performance, organize the inconsistent parts in the classification results. Use data recording tools to store these parts, obtaining the final performance analysis archive.

[0018] Based on a standardized dataset, a support vector machine (SVM) is used to classify the relationship between abrasion resistance and flexibility in water-based varnish samples. This aims to determine the boundary conditions of performance conflicts and identify potential conflict areas. This can be achieved using the following information technology methods. Assume a standardized dataset containing 80 samples is available. Abrasion resistance data is represented by standardized friction counts, ranging from -2.5 to 2.5, and flexibility data is represented by standardized bending angles, ranging from -2.0 to 2.0. First, the SVM classification algorithm is loaded using the Scikit-learn library in Python. The kernel function is set to radial basis function (RBF), and the dataset is divided into a training set (60 samples) and a test set (20 samples). Next, the SVM model parameters are optimized using a grid search method. For example, the regularization parameter C is set to 1.0, and the kernel function parameter gamma is set to 0.1 to ensure the model has a good fit to complex boundaries. Then, a classification model was built based on the training set, using abrasion resistance and flexibility data as input features. The classification labels were set as "high conflict" and "low conflict," with high conflict areas defined as samples with abrasion resistance values ​​greater than 1.5 and flexibility values ​​less than -1.0. A total of 15 samples were labeled as high conflict. After model training, the model was validated on the test set, and the classification accuracy was calculated. Assuming a result of 85%, this indicates that the model can effectively distinguish conflicting performance areas. Subsequently, the model was used to predict the classification results for the entire dataset, identifying potential conflict areas. For example, 10 samples were found to fall near the high conflict boundary, one of which had an abrasion resistance value of 1.8 and a flexibility value of -1.2, and was classified as high conflict. Finally, the classification results were combined with the original data to generate a new dataset containing sample numbers, standardized feature values, and conflict labels, stored as a CSV file for subsequent formulation adjustment analysis targeting conflict areas. This process, from classification modeling to conflict identification, forms a complete data analysis chain, providing technical support for optimizing the performance of water-based varnishes.

[0019] S103. Extract key feature vectors from the boundary conditions of performance conflicts, and mark them as high conflict points if the Euclidean distance of the feature vectors exceeds a preset threshold. Obtain a set of high conflict points as the basis for optimization.

[0020] Relevant data on performance conflicts are obtained from boundary conditions. Core data is extracted from this data to obtain a set of feature vectors. For this set of feature vectors, Euclidean distance is used to calculate the spatial difference between vectors. If the spatial difference exceeds a preset threshold, conflict detection is performed, high-conflict points are marked, and a preliminary list of high-conflict points is constructed. By marking the high-conflict point list, categorized conflict point groups are obtained, and their distribution characteristics are determined. Based on the conflict point grouping results, a set is constructed to obtain a set of high-conflict points. For this set of high-conflict points, correlation data with the optimization basis is obtained to determine the direction of subsequent processing.

[0021] When dealing with boundary conditions that conflict with performance, we first extract key feature vectors from the system through data preprocessing. Suppose we obtain data from a network performance optimization scenario, with boundary conditions including latency, throughput, and packet loss rate. We collect 100 sets of data for each, with latency ranging from 10ms to 200ms, throughput from 1Mbps to 50Mbps, and packet loss rate from 0.1% to 5%. We standardize these three indicators using the z-score method, transforming the data into a distribution with a mean of 0 and a standard deviation of 1. For example, the original values ​​of a set of data are 50ms latency, 20Mbps throughput, and 1% packet loss rate; after standardization, the resulting feature vector is [0.5, 0.3, -0.2]. Next, the Euclidean distance between feature vectors is calculated to determine the degree of conflict. A preset threshold of 2.0 is set. The algorithm calculates the distance for each pair of feature vectors. For example, the Euclidean distance between vector A[0.5,0.3,-0.2] and vector B[1.5,1.0,0.8] is sqrt((1.5-0.5)^2+(1.0-0.3)^2+(0.8-(-0.2))^2)=sqrt( The distance between vector A and vector C [2.5, 1.5, 1.5] is sqrt((2.5-0.5)^2+(1.5-0.3)^2+(1.5-(-0.2))^2)=sqrt(4+1.44+2.89)=2.86, which is greater than the threshold of 2.0, and is therefore marked as a high-collision point. The system automatically traverses all vector pairs, filters out combinations with a distance greater than 2.0, and finally obtains a set of high-collision points. For example, a total of 10 high-collision point pairs were identified, involving 15 feature vectors. These points are mainly concentrated in areas with a latency of more than 100ms and a packet loss rate of more than 3%. Based on this dataset, the system further analyzed the causes of conflicts and found a positive correlation between high latency and high packet loss rate, with a correlation coefficient of 0.75. This indicates that network congestion is likely the primary cause of the conflict, providing data support for subsequent optimization, such as prioritizing the adjustment of congestion control algorithm parameters. The entire process utilizes automated scripts to perform feature extraction, distance calculation, and conflict marking. Logically, it forms a closed loop from data collection to conflict identification and cause analysis, ensuring the accuracy of the optimization foundation.

[0022] S104. For a set of high-conflict points, a genetic algorithm is applied to generate an initial formula design scheme population, and multiple alternative formula design schemes are evolved through crossover and mutation operations.

[0023] For a set of high-conflict points, its internal set data is acquired and processed through data classification to obtain structured conflict data groups. Based on these structured conflict data groups, a genetic algorithm is used to generate a corresponding initial population, determining a preliminary formulation design framework. For this preliminary framework, crossover operations are performed on the individual data within the initial population to obtain a recombined set of design individuals. Based on this recombined set, mutation operations are performed, introducing a random adjustment mechanism to obtain diverse combinations of evolving individuals. For these diverse combinations, intermediate state data during the evolutionary process is acquired, and an optimized list of alternative solutions is determined through iterative screening. Based on this list, the matching degree data between the optimized alternative solutions and the formulation design is obtained. If the matching degree is lower than a preset threshold, further evolutionary processing is performed to obtain the final design population result. Finally, the final design population result is integrated through a solution generation module to obtain a complete set of formulation design solutions.

[0024] In network performance optimization scenarios, for a set of identified high-conflict points, the system automatically applies a genetic algorithm to generate an initial population of recipe design schemes, and evolves multiple alternative schemes through crossover and mutation operations. First, the system extracts 15 key data points from the high-conflict point set and generates an initial population based on the network parameters associated with each data point (such as bandwidth allocation ratio, buffer size, and retransmission count). Assuming the initial population contains 50 recipe schemes, each scheme has a bandwidth allocation ratio ranging from 0.2 to 0.8, a buffer size ranging from 2MB to 10MB, and a retransmission count ranging from 1 to 5. For example, an initial scheme might be [0.5, 6.0, 3]. Subsequently, the system uses the crossover operation of the genetic algorithm to pair the 50 schemes together, exchanging parameters using a single-point crossover method. For example, scheme A [0.5, 6.0, 3] is cross-crossed with scheme B [0.3, 8.0, 2] to generate new schemes [0.5, 8.0, 2] and [0.3, 6.0, 3]. Next, the system performs a mutation operation on the newly generated schemes, setting the mutation probability to 0.1 and randomly adjusting parameter values. For example, the bandwidth allocation ratio of a certain scheme is adjusted from 0.5 to 0.55, and the cache size is adjusted from 6.0 to 6.2, generating the mutated schemes [0.55, 6.2, 3]. Then, the system evaluates the performance of each scheme using a fitness function. Assuming the fitness function comprehensively considers network throughput improvement and latency reduction, the calculation formula is: Fitness = 0.6 * Throughput Gain - 0.4 * Latency Increase. The top 20% of schemes by fitness are selected for the next iteration, undergoing 10 iterations in total, ultimately generating 10 candidate scheme designs. The entire process is implemented through an automated algorithm, logically forming a closed loop from initial population generation to crossover mutation and then to fitness evaluation, ensuring the diversity of schemes and the rationality of the optimization direction. To further improve the schemes, the system matches and analyzes the candidate schemes with historical network load data, automatically adjusting parameter boundaries to form configuration combinations that better fit the actual scenario.

[0025] S105. The wear resistance and flexibility of the alternative formulation design schemes are evaluated through simulation tests, and the comprehensive score vector of each scheme is obtained to quantify the degree of performance balance.

[0026] Data was collected from alternative schemes in the formulation design using a simulation testing platform to obtain raw test data on wear resistance and flexibility for each scheme. Based on the collected raw test data, a standardization method was used to normalize the wear resistance and flexibility data, obtaining performance values ​​under a unified dimension. For the normalized performance values, a pre-established weighted allocation model was used to calculate the weighted results of wear resistance and flexibility, determining the comprehensive performance score for each scheme. If the comprehensive performance score was lower than a preset threshold, a secondary analysis was performed on the performance value of that scheme to obtain its deviation data in wear resistance and flexibility. Based on the deviation data, a support vector machine algorithm was used to classify the performance distribution of the alternative schemes, determining the classification of each scheme in terms of performance balance. Based on the classification results and the comprehensive performance score, a score vector for each scheme was constructed to obtain quantified performance balance level data. Based on the quantified performance balance level data, a ranking result for each alternative scheme was generated, determining the priority sequence of the optimal formulation design scheme.

[0027] When evaluating the wear resistance and flexibility of alternative formulation designs, performance data was first generated through simulation tests. Assuming three formulations, A, B, and C, virtual wear and bending tests were conducted, yielding wear resistance data (unit: wear rate, %) of A: 3.5, B: 2.8, and C: 4.2, and flexibility data (unit: bending angle, degrees) of A: 75, B: 60, and C: 80. Next, a standardization algorithm was used to process the data, mapping the wear resistance and flexibility data to a range of 0 to 1. The calculation formula is: Standardized value = (Original value - Minimum value) / (Maximum value - Minimum value). Inverse standardization was used for wear resistance to reflect high performance with low wear rates. The results were: A: Wear resistance: 0.41, Flexibility: 0.75; B: Wear resistance: 0.82, Flexibility: 0.0; C: Wear resistance: 0.0, Flexibility: 1.0. Subsequently, a weighted average algorithm was used to calculate the comprehensive score vector. Assuming abrasion resistance had a weight of 0.6 and flexibility a weight of 0.4, the calculation formula was: Comprehensive Score = Abrasion Resistance Standardized Value × 0.6 + Flexibility Standardized Value × 0.4, yielding scores of A: 0.546, B: 0.492, and C: 0.4. Finally, the performance balance was analyzed by comparing the comprehensive score vectors. A balance threshold of 0.5 was set. Scheme A scored above the threshold, and the difference between the abrasion resistance and flexibility standardized values ​​was only 0.34, indicating relatively balanced performance. Scheme B showed outstanding abrasion resistance but poor flexibility (difference of 0.82), while Scheme C exhibited excellent flexibility but insufficient abrasion resistance (difference of 1.0). Comprehensive analysis concluded that Scheme A was the optimal choice in terms of performance balance. All of the above processes can be automated through programming. After data input, the system completes standardization, weighted scoring, and balance assessment, and outputs a visual report to assist decision-making. It can also be linked to business needs such as cost analysis, combining the score vector with cost data to further optimize the formulation selection logic.

[0028] S106. If the average value of the comprehensive score vector is lower than the preset threshold, the genetic algorithm is iterated to adjust the formula design scheme; otherwise, the scheme with the highest score is selected as the preliminary optimization result.

[0029] By acquiring comprehensive scoring data from multiple design schemes, scoring vectors are constructed for subsequent analysis, resulting in an initial set of scoring vectors. The mean of these vectors is calculated and standardized using statistical tools to determine if it meets a preset threshold. If the mean is lower than the threshold, a genetic algorithm is triggered to iteratively adjust the formulation design, generating a new set of design schemes. For this new set, comprehensive scoring data is reacquired, and updated scoring vectors are constructed, resulting in a new set of scoring vectors. If the updated mean of the scoring vectors is still lower than the threshold, the genetic algorithm continues to adjust the schemes, iterating until the mean reaches or exceeds the threshold. When the mean of the scoring vectors reaches or exceeds the threshold, the scheme with the highest score is selected as the initial optimization result, and the final design scheme is output.

[0030] This section discusses the technical aspects of determining the average value of the comprehensive score vector and subsequent iterative adjustment of the formula design scheme using a genetic algorithm. Assume we have a formula design system to optimize the proportion of a certain food additive. The score vector consists of three dimensions: sweetness, taste, and cost, with each dimension having a maximum score of 10 points. First, the system collects 10 initial formula schemes and calculates the comprehensive score vector for each scheme. For example, the score vector for scheme 1 is (8.5, 7.0, 6.5). Using a weighted average (weights of 0.4, 0.3, and 0.3 respectively), the comprehensive score is calculated as 8.5 × 0.4 + 7.0 × 0.3 + 6.5 × 0.3 = 7.45. Similarly, after calculating the comprehensive scores for all 10 schemes, a score list is obtained: [7.45, 6.8, 7.1, 6.9, 7.3, 6.5, 7.0, 6.7, 7.2, 6.6], with an average value of 6.95. A preset threshold of 7.5 is set. Since 6.95 is lower than 7.5, the system automatically triggers iterative optimization using a genetic algorithm. The genetic algorithm first selects the two highest-scoring schemes (7.45 and 7.3) as parents. A new formula is generated through crossover operations; for example, the sweetness parameter of scheme 1 is combined with the taste parameter of scheme 5, while random mutation is introduced (e.g., increasing the sweetness parameter by 0.1), generating a new scheme score vector (8.6, 7.2, 6.4) with a comprehensive score of 7.58. After 10 iterations, the average score of the new generation of schemes increases to 7.6, exceeding the threshold of 7.5. The system stops iterating and selects the scheme with the highest comprehensive score (8.2) as the initial optimization result, outputting its formula parameters (e.g., 3.2% sweetener, 1.5% taste modifier, and cost controlled at 0.8 yuan per unit). The entire process is completed automatically by the system. The update of the score vector and the crossover mutation of the genetic algorithm are driven by preset rules, ensuring rigorous logic and traceable results, while also being closely related to the business needs of food formula optimization.

[0031] S107. Obtain material stability indicators from the preliminary optimization results, and determine the final formulation design scheme by comparing it with the original water-based varnish data to achieve an effective balance between wear resistance and flexibility.

[0032] Step 1: Obtain material stability index data from the preliminary optimization results. Use data extraction techniques to classify and organize the various indicators to obtain a structured index dataset. Step 2: Compare and analyze the structured index dataset with the original water-based varnish data. Identify the differences between the two using data matching methods to determine the difference dataset. Step 3: Based on the difference dataset, use a support vector machine algorithm to classify and predict the relevant indicators of wear resistance and flexibility, obtaining a preliminary evaluation result of performance balance. Step 4: If the wear resistance index is higher than the flexibility index in the preliminary evaluation result, adjust the weights of the formulation parameters and recalculate the performance balance value to obtain adjusted evaluation data. Step 5: Using the adjusted evaluation data and the constraints of the final formulation design scheme, generate multiple alternative formulation combinations to determine the alternative formulation list. Step 6: For the alternative formulation list, obtain simulated test data for each formulation group, analyze its comprehensive performance in wear resistance and flexibility, and determine the optimal formulation scheme. Step 7: Based on the optimal formulation scheme, generate the final formulation design output data, completing the performance balance index evaluation process.

[0033] In the process of obtaining material stability indicators and determining the final formulation design, key data, such as wear resistance and flexibility, are first extracted from the preliminary optimization results using information technology. For example, the preliminary optimization results show that a certain formulation has a wear resistance value of 85 (based on standard wear tests, in cycles) and a flexibility value of 7.5 (based on bending tests, in millimeters of bending radius). The system uses a database comparison algorithm to compare these data with the original water-based varnish data, which has a wear resistance value of 70 and a flexibility value of 8.0. By calculating the percentage difference, the wear resistance improved by 21.4% (calculation formula: (85-70) / 70*100), while the flexibility decreased by 6.25% (calculation formula: (8.0-7.5) / 8.0*100). Subsequently, based on a preset balance model, the system sets the weight ratio of wear resistance to flexibility at 6:4 and calculates the comprehensive performance score. The initial formula score is 85*0.6 + 7.5*0.4 = 54, while the original data score is 70*0.6 + 8.0*0.4 = 45.2, indicating that the initial formula has better overall performance. Next, the system further adjusts the formula parameters through a multi-objective optimization algorithm. For example, increasing the proportion of flexibility additive by 0.5% predicts that the flexibility value will increase to 7.8, while the wear resistance will slightly decrease to 83. The recalculated comprehensive score is 83*0.6 + 7.8*0.4 = 52.9, which is still higher than the original data, ensuring balance. Finally, the system generates a formula design scheme, outputs the adjusted formula parameters, and stores the results in the database. It also links to the subsequent production process parameter adjustment module to ensure automated integration with the production line, forming a complete logical chain from data analysis to scheme output. Through the above process, the system not only achieves performance balance but also provides data support for subsequent process optimization.

[0034] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A comprehensive manufacturing method for optimizing the performance of water-based abrasion-resistant plastic substitute varnish, characterized in that, include: Abrasion resistance and flexibility data of coating samples are obtained, and noise is removed through preprocessing to obtain a standardized dataset. Based on the standardized dataset, a support vector machine is used to classify the relationship between abrasion resistance and flexibility, determining the boundary conditions of performance conflicts. Feature vectors are extracted from the boundary conditions of performance conflicts, and high-conflict points are identified using Euclidean distance to obtain a set of high-conflict points. For the set of high-conflict points, a genetic algorithm is used to generate and evolve alternative formulation design schemes. The abrasion resistance and flexibility of the alternative formulation design schemes are evaluated through simulation testing to obtain a comprehensive score vector. The relationship between the average value of the comprehensive score vector and a preset threshold determines whether to iterate the genetic algorithm; otherwise, the highest-scoring scheme is selected as the preliminary optimization result. Material stability indicators are obtained from the preliminary optimization results, and the final formulation design scheme is determined by comparing them with the original coating data.

2. The comprehensive manufacturing method for optimizing the performance of water-based abrasion-resistant plastic substitute varnish as described in claim 1, characterized in that, The process of obtaining abrasion resistance and flexibility data from coating samples and preprocessing to remove noise to obtain a standardized dataset includes: constructing an initial dataset by obtaining raw data of abrasion resistance and flexibility values ​​from coating samples; if the noise in the initial dataset exceeds a preset threshold, preprocessing is performed using a filtering method to obtain a denoised dataset; standardizing the abrasion resistance and flexibility values ​​based on the denoised dataset to generate a standardized dataset; if the performance indicators in the standardized dataset meet preset conditions, recording quality assessment parameters to obtain assessment criteria; classifying the abrasion resistance and flexibility values ​​in the standardized dataset using a support vector machine based on the assessment criteria to determine the classification results; constructing an analysis model based on the classification results to obtain performance prediction output data; and adjusting the analysis model parameters to determine the performance distribution if there is a deviation between the performance prediction output data and the quality assessment parameters.

3. The comprehensive manufacturing method for optimizing the performance of water-based abrasion-resistant plastic substitute varnish as described in claim 1, characterized in that, The process of classifying the relationship between wear resistance and flexibility using support vector machines (SVM) based on the standardized dataset and determining the boundary conditions of performance conflicts includes: obtaining records of wear resistance and flexibility from the standardized dataset; performing preliminary screening using data processing methods to obtain a set of organized performance indicators; classifying wear resistance and flexibility separately based on the organized performance indicator set; using SVM to divide the interaction between the two to determine preliminary classification results; analyzing the conflict boundaries in the performance indicators through the preliminary classification results to identify inconsistencies between wear resistance and flexibility and obtain the distribution range of conflict boundaries; performing data processing on potential conflict areas based on the distribution range of conflict boundaries; marking areas where performance indicators deviate from preset thresholds to determine marked conflict areas; obtaining key points from the marked conflict areas; conducting in-depth comparisons of these points to determine the conflicting performance of wear resistance and flexibility under specific conditions; and storing the inconsistent parts in the classification results based on the conflicting performance to obtain a performance analysis archive using data recording tools.

4. The comprehensive manufacturing method for optimizing the performance of water-based abrasion-resistant plastic substitute varnish as described in claim 1, characterized in that, The step of extracting feature vectors from the boundary conditions of the performance conflict and obtaining a set of high conflict points by judging high conflict points through Euclidean distance includes: obtaining relevant data on performance conflicts from the boundary conditions; performing core extraction on the data to obtain a set of feature vectors; calculating the spatial difference value between each vector using the Euclidean distance method on the set of feature vectors; if the spatial difference value exceeds a preset threshold, performing conflict judgment to mark high conflict points and constructing a preliminary list of high conflict points; obtaining classified conflict point groups by marking the preliminary list of high conflict points and judging their distribution characteristics; performing set construction based on the conflict point grouping results to obtain a set of high conflict points; and obtaining the correlation data between the high conflict point set and the optimization basis to determine the direction of subsequent processing.

5. The comprehensive manufacturing method for optimizing the performance of water-based abrasion-resistant plastic substitute varnish as described in claim 1, characterized in that, The step of generating and evolving alternative formulation design schemes using a genetic algorithm for the set of high conflict points includes: acquiring internal set data for the set of high conflict points and processing it through data classification to obtain structured conflict data groups; generating an initial population based on the structured conflict data groups using a genetic algorithm to determine a preliminary formulation design framework; processing individual data in the initial population through crossover operations to obtain a recombined design individual set based on the preliminary formulation design framework; performing mutation operations on the recombined design individual set to introduce a random adjustment mechanism to obtain diverse evolutionary individual combinations; acquiring intermediate state data during the evolution process from the diverse evolutionary individual combinations and determining an optimized alternative scheme list through iterative screening; obtaining matching degree data between the optimized alternative scheme list and the formulation design; if the matching degree is lower than a preset threshold, performing further scheme evolution processing to obtain the final design population result; and integrating the final design population result through a scheme generation module to obtain a set of formulation design schemes.

6. The comprehensive manufacturing method for optimizing the performance of water-based abrasion-resistant plastic substitute varnish as described in claim 1, characterized in that, The step of evaluating the wear resistance and flexibility of the alternative formulation design schemes through simulation testing to obtain a comprehensive score vector includes: acquiring raw test data on wear resistance and flexibility of each scheme through a simulation testing platform; normalizing the wear resistance and flexibility data using a standardization method to obtain performance values ​​under a unified dimension; calculating the weighted results of wear resistance and flexibility using a pre-established weighting model to determine the comprehensive performance score of each scheme; if the comprehensive performance score is lower than a preset threshold, performing a secondary analysis on the performance value of the scheme to obtain its deviation data in wear resistance and flexibility; classifying the performance distribution of the alternative schemes using a support vector machine algorithm based on the deviation data to determine the classification of each scheme in terms of performance balance; constructing a score vector for each scheme by combining the classification results with the comprehensive performance score to obtain quantified performance balance level data; and generating a ranking result for each alternative scheme based on the quantified performance balance level data to determine the priority sequence of the optimal formulation design scheme.

7. The comprehensive manufacturing method for optimizing the performance of water-based abrasion-resistant plastic substitute varnish as described in claim 1, characterized in that, The step of determining whether to iterate the genetic algorithm based on the relationship between the average value of the comprehensive score vector and a preset threshold, and otherwise selecting the highest-scoring scheme as the preliminary optimization result, includes: obtaining comprehensive score data from multiple design schemes to construct a score vector set; calculating the mean of the vectors based on the initial score vector set; standardizing the mean using statistical tools to determine whether the mean reaches a preset threshold; if the mean of the vectors is lower than the preset threshold, triggering the genetic algorithm to iteratively adjust the formula design to generate a new set of design schemes; obtaining comprehensive score data again for the new set of design schemes to construct an updated score vector set; if the mean of the updated score vectors is still lower than the preset threshold, continuing to adjust the schemes through the genetic algorithm until the mean reaches or exceeds the preset threshold; and when the mean of the score vectors reaches or exceeds the preset threshold, selecting the scheme with the highest score as the preliminary optimization result and outputting the design scheme.

8. The comprehensive manufacturing method for optimizing the performance of water-based abrasion-resistant plastic substitute varnish as described in claim 1, characterized in that, The process of obtaining material stability indicators from the preliminary optimization results and determining the final formulation design scheme by comparing them with the original coating data includes: obtaining material stability indicator data from the preliminary optimization results; classifying and organizing the various indicators using data extraction techniques to obtain a structured indicator dataset; comparing and analyzing the structured indicator dataset with the original coating data; identifying the differences between the two using data matching methods to determine the difference dataset; using the support vector machine algorithm to classify and predict the relevant indicators of wear resistance and flexibility based on the difference dataset to obtain a preliminary evaluation result of performance balance; if the wear resistance index is higher than the flexibility index in the preliminary evaluation result, adjusting the weights of the formulation parameters and recalculating the performance balance value to obtain adjusted evaluation data; generating multiple alternative formulation combinations based on the adjusted evaluation data and the constraints of the final formulation design scheme to determine an alternative formulation list; obtaining simulated test data for each formulation in the alternative formulation list to analyze its comprehensive performance in wear resistance and flexibility to determine the optimal formulation scheme; and generating the final formulation design output data based on the optimal formulation scheme to complete the performance balance indicator evaluation process.

9. The comprehensive manufacturing method for optimizing the performance of water-based abrasion-resistant plastic substitute varnish as described in claim 1, characterized in that, The process of generating and evolving alternative formulation design schemes using a genetic algorithm and evaluating the wear resistance and flexibility of the alternative formulation design schemes through simulation testing to obtain a comprehensive score vector includes: generating an initial population using a genetic algorithm for a set of high conflict points, and then evolving diverse alternative formulation design schemes through crossover and mutation operations; obtaining raw test data for the alternative formulation design schemes through simulation testing and performing normalization to obtain performance values; calculating a weighted comprehensive performance score based on the performance values ​​and constructing a score vector; if the comprehensive performance score or the average value of the score vector is lower than a preset threshold, iterating the genetic algorithm to adjust the formulation design scheme until the threshold condition is met; determining the highest-scoring scheme through the score vector as the preliminary optimization result, and further comparing the material stability index with the original coating data to determine the final formulation design scheme.

Images