Quality rating and recommending method for magnesium-based alloy

Abstract

The invention relates to the technical field of magnesium-based alloy manufacturing data processing, provides a magnesium-based alloy quality rating and recommending method, and aims to solve the problem of inaccurate quality rating caused by raw material component fluctuation and test condition difference in the existing magnesium-based alloy manufacturing process. The method comprises the following steps: constructing a preprocessing data set and performing feature extraction and integration by obtaining magnesium-based alloy preparation historical tasks, raw material components, performance test, test conditions, test equipment information and operator skill level data; and training and establishing a quality rating model by using a neural network model, and identifying the influence of raw material components and test conditions on the alloy quality. And through simulation and optimization, preparation parameters matched with the real-time magnesium-based alloy quality grade are obtained and serve as recommended parameters. According to the method, the accuracy of magnesium-based alloy quality rating is improved, and a magnesium-based alloy recommendation method is provided.

Description

Technical Field

[0001] The invention relates to the technical field of magnesium-based alloy production data processing, and in particular to a quality rating and recommendation method for magnesium-based alloys. Background Art

[0002] Magnesium-based alloys are alloys made of magnesium as the base metal and aluminum, zinc, manganese, rare earth, titanium and other alloying elements. Magnesium-based alloys have excellent specific strength and specific stiffness. Magnesium-based alloys are a lightweight, high-strength structural material and are widely used in aerospace, military, transportation, electronics, chemical and other fields. For example, in the aerospace industry, magnesium-based alloys are widely used in the manufacture of aircraft engine parts, bracket structures, etc. due to their light weight, good damping performance and vibration resistance; in the electronics field, magnesium-based alloys have become ideal materials for the manufacture of 3C product shells such as mobile phones, digital cameras, and laptops due to their excellent thin-wall casting performance, good specific strength and anti-collision ability.

[0003] After the quality rating of magnesium-based alloys, necessary performance tests are carried out, such as tensile test, hardness test, corrosion resistance test, etc., to verify whether the actual performance of magnesium-based alloys meets the design requirements. For magnesium-based alloy parts, fatigue life assessment and environmental adaptability tests are also carried out to simulate the complex working conditions in actual use. Based on the test results, targeted recommendations are provided for the application of magnesium-based alloys.

[0004] The quality rating of existing magnesium-based alloys is mainly based on a comprehensive evaluation of multiple indicators such as their chemical composition, mechanical properties, physical properties, and processing technology. The recommended method focuses on selecting alloy grades that meet application requirements, strictly controlling quality parameters during the production process, and conducting necessary performance tests and verifications. However, in terms of data processing, fluctuations in the composition of raw materials from different sources also affect the quality of the final magnesium-based alloy. Due to differences in test equipment, test conditions (such as temperature, humidity, loading rate, etc.) and operator skill levels, the performance test data of the same batch of alloys fluctuates, affecting the accuracy of the quality rating. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention provides a quality rating and recommendation method for magnesium-based alloys to solve the problem that fluctuations in the composition of raw materials from different sources also affect the quality of the final magnesium-based alloys. Due to differences in testing equipment, testing conditions and operator skill levels, performance test data of the same batch of alloys fluctuate, affecting the accuracy of quality rating.

[0006] In order to solve the above technical problems, the specific technical solutions of the present invention are as follows: The present invention provides a quality rating and recommendation method for a magnesium-based alloy, comprising: Step S101, obtaining magnesium-based alloy preparation history tasks, magnesium-based alloy raw material composition data, magnesium-based alloy performance test data, magnesium-based alloy test condition data, magnesium-based alloy test equipment information, and magnesium-based alloy preparation operator skill level data to obtain a magnesium-based alloy data set, and preprocessing the magnesium-based alloy data set to obtain a preprocessed magnesium-based alloy data set; Step S102, extracting features from the preprocessed magnesium-based alloy data set to obtain magnesium-based alloy preparation history task features and magnesium-based alloy data features, the magnesium-based alloy data features include magnesium-based alloy data raw material composition features, magnesium-based alloy data performance test features, magnesium-based alloy data test condition features, magnesium-based alloy data test equipment features, and magnesium-based alloy data operator skill level features, and integrating the magnesium-based alloy data features to obtain magnesium-based alloy data statistical analysis results; Step S103, based on the magnesium-based alloy data characteristics and the magnesium-based alloy data statistical analysis results, a neural network model is used for training to establish a quality rating model for the magnesium-based alloy, where the quality rating model for the magnesium-based alloy is used to identify the relationship between raw material composition, test conditions and alloy quality; Step S104, obtaining historical data of magnesium-based alloy preparation, substituting the historical data of magnesium-based alloy preparation into a quality rating model of magnesium-based alloy, and obtaining a quality data set of magnesium-based alloy; Step S105, collecting raw material composition data in the magnesium-based alloy to be analyzed, matching the raw material composition data in the magnesium-based alloy to be analyzed with the quality data set of the magnesium-based alloy, obtaining the magnesium-based alloy quality level corresponding to the raw material composition data in the magnesium-based alloy to be analyzed, and using the magnesium-based alloy quality level corresponding to the raw material composition data in the magnesium-based alloy to be analyzed as the real-time magnesium-based alloy quality level; Step S106, calling up the magnesium-based alloy preparation task corresponding to the real-time magnesium-based alloy quality level, taking the magnesium-based alloy preparation task corresponding to the real-time magnesium-based alloy quality level as the magnesium-based alloy preparation task to be simulated, substituting the magnesium-based alloy preparation task to be simulated into a preset magnesium-based alloy preparation simulation model, obtaining simulation parameters of the magnesium-based alloy preparation task to be simulated, comparing the simulation parameters of the magnesium-based alloy preparation task to be simulated with the real-time magnesium-based alloy quality level, and obtaining magnesium-based alloy preparation parameter error data; Step S107, matching the magnesium-based alloy preparation error data with a preset magnesium-based alloy preparation error range knowledge base to obtain a magnesium-based alloy preparation error result. If the magnesium-based alloy preparation error result exceeds the standard error value, then retrieving the data exceeding the standard error value in the magnesium-based alloy preparation error data, marking the data exceeding the standard error value as a first type of label, marking the data not exceeding the standard error value as a second type of label, and aggregating the data corresponding to the first type of label and the second type of label to obtain a dynamic label data set; Step S108, substitute the dynamic tag data set into the preset magnesium-based alloy preparation model, output magnesium-based alloy preparation simulation data, optimize the magnesium-based alloy preparation simulation data based on the use of the predator optimization algorithm of the preset magnesium-based alloy preparation model to obtain optimized magnesium-based alloy preparation simulation data, and perform data processing on the optimized magnesium-based alloy preparation simulation data through the preset magnesium-based alloy production model to obtain optimized magnesium-based alloy preparation simulation parameters, match the optimized magnesium-based alloy preparation simulation parameters with the data corresponding to the real-time magnesium-based alloy quality level, obtain magnesium-based alloy preparation parameter data matching the real-time magnesium-based alloy quality level, and use the magnesium-based alloy preparation parameter data matching the real-time magnesium-based alloy quality level as recommended parameters; Step S109, matching the recommended parameters with the historical data of magnesium-based alloy preparation, and using the magnesium-based alloy prepared by using the recommended parameters as the recommended magnesium-based alloy.

[0007] Furthermore, the quality rating and recommendation method of magnesium-based alloys according to the present invention, step S102, comprises: The characteristics of the magnesium-based alloy preparation history task include the magnesium-based alloy preparation task ID, magnesium-based alloy preparation date, magnesium-based alloy preparation batch, and magnesium-based alloy preparation process parameters; The raw material composition characteristics of magnesium-based alloy data include the content characteristics of each element of magnesium-based alloy data; The data performance test characteristics of magnesium-based alloys include magnesium-based alloy tensile strength characteristics, magnesium-based alloy yield strength characteristics, magnesium-based alloy elongation characteristics and magnesium-based alloy hardness characteristics; The test condition characteristics of magnesium-based alloy data include the test temperature characteristics of magnesium-based alloy data, the test humidity characteristics of magnesium-based alloy data, and the test loading rate characteristics of magnesium-based alloy data; The characteristics of the magnesium-based alloy data testing equipment include the characteristics of the magnesium-based alloy data testing equipment model, the characteristics of the magnesium-based alloy data testing specification, and the characteristics of the magnesium-based alloy data testing calibration record; The skill level characteristics of magnesium-based alloy data operators include the training record characteristics of magnesium-based alloy data operators, the years of experience characteristics of magnesium-based alloy data operators, and the historical operation accuracy characteristics of magnesium-based alloy data operators; The statistical analysis results of magnesium-based alloy data are obtained by integrating the characteristics of historical tasks in magnesium-based alloy preparation, raw material composition characteristics of magnesium-based alloy data, performance test characteristics of magnesium-based alloy data, test condition characteristics of magnesium-based alloy data, test equipment characteristics of magnesium-based alloy data and skill level characteristics of magnesium-based alloy data operators.

[0008] Furthermore, the quality rating and recommendation method of magnesium-based alloys described in the present invention, step S103, comprises: Using the statistical analysis results of magnesium-based alloy data as the input of the neural network model, initializing the parameters of the neural network model, and obtaining the initialized neural network model; The statistical analysis results of magnesium-based alloy data are divided into a training set and a validation set. The initialized neural network model is trained using the training set. The output is calculated through forward propagation, the error is calculated according to the loss function, and the model parameters are updated through back propagation. The training process is continuously iterated until the preset training rounds are reached or the performance on the validation set is no longer improved, then the trained neural network model is output; The trained neural network model is encapsulated as a quality rating model for magnesium-based alloys. The raw material composition data of magnesium-based alloys, performance test data of magnesium-based alloys, test condition data of magnesium-based alloys, test equipment information of magnesium-based alloys, and skill level data of magnesium-based alloy preparation operators are used as input, and the rating results of the quality of magnesium-based alloys are used as output.

[0009] Furthermore, the quality rating and recommendation method of magnesium-based alloys described in the present invention, step S104, comprises: Obtaining historical preparation data of magnesium-based alloys from a preset database, the historical preparation data of magnesium-based alloys including historical magnesium-based alloy raw material composition, historical magnesium-based alloy test conditions, historical magnesium-based alloy test equipment information, and historical magnesium-based alloy operator records during the preparation process; Preprocessing the preparation history data of the magnesium-based alloy to obtain the preprocessed preparation history data of the magnesium-based alloy; According to the feature requirements of the magnesium-based alloy quality rating model, the features required by the magnesium-based alloy quality rating model are extracted from the pre-processed magnesium-based alloy preparation history data to obtain an extracted magnesium-based alloy data feature set. The extracted magnesium-based alloy data feature set includes the features required by the magnesium-based alloy quality rating model, including magnesium-based alloy raw material composition data features, magnesium-based alloy performance test data features, magnesium-based alloy test condition data features, magnesium-based alloy test equipment information features, and magnesium-based alloy preparation operator skill level data features; The extracted magnesium-based alloy data feature set is input into the magnesium-based alloy quality rating model. The magnesium-based alloy quality rating model calculates the quality rating of the magnesium-based alloy through the neural network structure inside the magnesium-based alloy quality rating model according to the input data. The quality rating data of the magnesium-based alloy includes the quality score of the magnesium-based alloy, the quality grade of the magnesium-based alloy and the quality classification result of the magnesium-based alloy; The quality rating data of the magnesium-based alloy output by the magnesium-based alloy quality rating model is sorted out, and the quality data of the magnesium-based alloy is associated with the original magnesium-based alloy preparation history data to obtain a quality data set of the magnesium-based alloy.

[0010] Furthermore, the quality rating and recommendation method of magnesium-based alloys described in the present invention, step S105, comprises: Use magnesium-based alloy raw material composition testing equipment to test the magnesium-based alloy sample to be analyzed, and collect raw material composition data of the magnesium-based alloy to be analyzed; Preprocessing the raw material composition data of the magnesium-based alloy to be analyzed, matching the preprocessed raw material composition data of the magnesium-based alloy to be analyzed with the data in the magnesium-based alloy quality data set, and obtaining historical data records matching the raw material composition of the magnesium-based alloy to be analyzed based on a preset similarity calculation algorithm; Extracting historical data of quality level of magnesium-based alloy from historical data records matching the raw material composition of magnesium-based alloy to be analyzed, and obtaining the quality level of magnesium-based alloy corresponding to the raw material composition data in magnesium-based alloy to be analyzed; The magnesium-based alloy quality level corresponding to the raw material composition data in the magnesium-based alloy to be analyzed is used as the real-time magnesium-based alloy quality level of the magnesium-based alloy to be analyzed.

[0011] Furthermore, the quality rating and recommendation method of magnesium-based alloys described in the present invention, step S106, comprises: According to the real-time magnesium-based alloy quality level, query and retrieve the magnesium-based alloy preparation task record corresponding to the real-time magnesium-based alloy quality level in a preset database to obtain the magnesium-based alloy preparation task corresponding to the real-time magnesium-based alloy quality level; The magnesium-based alloy preparation task corresponding to the real-time magnesium-based alloy quality level is determined as the magnesium-based alloy preparation task to be simulated, and the magnesium-based alloy preparation task to be simulated is substituted into a preset magnesium-based alloy preparation simulation model, and the preset magnesium-based alloy preparation simulation model outputs simulation parameters of the magnesium-based alloy preparation task to be simulated, and the simulation parameters of the magnesium-based alloy preparation task to be simulated include the expected alloy properties of magnesium-based alloy preparation, variables in the magnesium-based alloy preparation process, control parameters of magnesium-based alloy preparation equipment, and content parameters of each component in magnesium-based alloy preparation; The simulation parameters of the magnesium-based alloy preparation task to be simulated are compared with the real-time magnesium-based alloy quality level to obtain the comparison result of the simulation parameters of the magnesium-based alloy preparation task to be simulated and the real-time magnesium-based alloy quality level, that is, the magnesium-based alloy preparation parameter error data. The magnesium-based alloy preparation parameter error data includes the magnesium-based alloy preparation error value, the magnesium-based alloy preparation error range and the magnesium-based alloy preparation error source analysis.

[0012] Furthermore, the quality rating and recommendation method of magnesium-based alloys according to the present invention, step S107, comprises: Matching each error index in the magnesium-based alloy preparation error data with the standard error range in the preset error range knowledge base one by one, and obtaining a matching result, i.e., a magnesium-based alloy preparation error result, wherein the magnesium-based alloy preparation error result includes each magnesium-based alloy preparation error index; Determine whether there are indicators exceeding the standard error value for each magnesium-based alloy preparation error index in the magnesium-based alloy preparation error result, and obtain magnesium-based alloy preparation error index data items exceeding the standard error value; Retrieving magnesium-based alloy preparation error index data items exceeding the standard error value from magnesium-based alloy preparation error data; The magnesium-based alloy preparation error index data items exceeding the standard error value are marked as first-class labels, and the magnesium-based alloy preparation error index data items not exceeding the standard error value are marked as second-class labels; The first category label data and the second category label data are aggregated to form a data set including the first category label data and the second category label data, and the data set including the first category label data and the second category label data is used as a dynamic label data set.

[0013] Furthermore, the quality rating and recommendation method of magnesium-based alloys described in the present invention, step S108, comprises: Substituting the data in the dynamic label data set into a preset magnesium-based alloy preparation model as input parameters of the magnesium-based alloy preparation model, the preset magnesium-based alloy preparation model preliminarily simulates the preparation results according to the input parameters to obtain magnesium-based alloy preparation simulation data; Apply the predatory optimization algorithm to optimize the simulation data of magnesium-based alloy preparation, and output the optimized simulation data of magnesium-based alloy preparation; Substituting the optimized magnesium-based alloy preparation simulation data into a preset magnesium-based alloy production model, and outputting the optimized magnesium-based alloy preparation simulation parameters; Matching the optimized magnesium-based alloy preparation simulation parameters with the data corresponding to the real-time magnesium-based alloy quality level to obtain magnesium-based alloy preparation parameter data matching the real-time magnesium-based alloy quality level; The magnesium-based alloy preparation parameter data matching the real-time magnesium-based alloy quality level is determined as the recommended parameters, and the recommended parameters are the process parameters for the preparation of the magnesium-based alloy.

[0014] Furthermore, the quality rating and recommendation method of magnesium-based alloys described in the present invention, step S109, comprises: According to the magnesium-based alloy preparation data, a similarity calculation model is constructed, and the characteristic vector of each data is extracted from the magnesium-based alloy preparation history data to obtain the characteristic vector of magnesium-based alloy preparation, and the characteristic vector of each data is extracted from the recommended parameters to obtain the characteristic vector of the recommended parameters; Using the constructed similarity calculation model, the feature vector of the recommended parameters is compared with the feature vector of the magnesium-based alloy preparation, and the similarity score between the feature vector of the recommended parameters and the feature vector of the magnesium-based alloy preparation is calculated; According to the similarity score between the characteristic vector of the recommended parameter and the characteristic vector of the magnesium-based alloy preparation, the historical data of the magnesium-based alloy preparation are sorted to obtain the sorting result of the historical data of the magnesium-based alloy preparation, and the data ranked first in the sorting result of the historical data of the magnesium-based alloy preparation is used as the data to be processed; The magnesium-based alloy finished product information corresponding to the data to be processed is retrieved from the magnesium-based alloy preparation history data, and the magnesium-based alloy finished product information corresponding to the data to be processed is used as the recommended magnesium-based alloy.

[0015] Furthermore, the quality rating and recommendation method of magnesium-based alloys according to the present invention, said step S108, further comprises: Receiving and setting the initial conditions of the predator optimization algorithm, the initial conditions of the predator optimization algorithm include initializing the population, determining the operating parameters of the algorithm, substituting each set of data in the initialized population into a preset magnesium-based alloy preparation model, calculating the simulation results corresponding to each set of data in the initialized population, and evaluating the fitness of each set of data according to a preset fitness function to obtain the fitness evaluation result of the initial population; According to the fitness evaluation results of the initial population, the individual with the highest fitness evaluation result of the initial population is selected as the prey, and the individual with the lowest fitness evaluation result of the initial population is selected as the predator. The predator updates its parameters by imitating the parameters of the prey and outputs the updated population data; Substituting the updated population data into the preset magnesium-based alloy preparation model again, calculating new simulation results, and evaluating the fitness of each group of data in the new simulation results to obtain the fitness evaluation results of the updated population; Check whether the fitness evaluation result of the updated population meets the termination condition. If so, output the optimization result. If not satisfied, continue to perform the predation operation; If the judgment result is to continue iteration, the population data will continue to be updated, and the fitness of the new population data will be evaluated. The population data and fitness evaluation results after each round of iteration will be output until the termination condition is met; The individual with the highest fitness is selected from the final population as the optimized simulation data for magnesium-based alloy preparation.

[0016] Beneficial effects of the present invention: The present invention comprehensively considers multiple factors such as raw material composition, test conditions, test equipment, operator skill level, etc., and uses a neural network model for training, so as to more accurately evaluate the quality of magnesium-based alloys. By simulating and optimizing preparation parameters, the present invention can reduce errors in the actual preparation process and improve the quality consistency of magnesium-based alloys. The automated quality rating and recommendation method provided by the present invention reduces the need for manual intervention, thereby improving production efficiency. The automated process not only speeds up the evaluation, but also reduces the possibility of human error. By optimizing preparation parameters and reducing scrap rates, the present invention helps to reduce production costs. More accurate quality ratings and preparation parameter optimization reduce waste caused by quality problems, thereby improving overall economic efficiency.

[0017] In summary, the present invention, through its unique technical solution, shows remarkable beneficial effects in improving the accuracy of quality rating, optimizing preparation parameters, improving production efficiency, reducing production costs and enhancing data traceability. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solution of the present invention, the following is a brief introduction to the drawings required for use in the embodiments. Obviously, for ordinary technicians in this field, other drawings can be obtained based on the drawings without paying any creative labor.

[0019] Figure 1 A schematic flow chart of a quality rating and recommendation method for magnesium-based alloys provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described in conjunction with the specific embodiments of the present invention and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. The technical solutions provided by the embodiments of the present invention are described below in conjunction with the drawings. In order to better understand the purpose of the present invention, the present invention is further described below.

[0021] The present invention provides a quality rating and recommendation method for a magnesium-based alloy, comprising: Step S101, obtaining magnesium-based alloy preparation history tasks, magnesium-based alloy raw material composition data, magnesium-based alloy performance test data, magnesium-based alloy test condition data, magnesium-based alloy test equipment information, and magnesium-based alloy preparation operator skill level data to obtain a magnesium-based alloy data set, and preprocessing the magnesium-based alloy data set to obtain a preprocessed magnesium-based alloy data set; Preparation history task data: Extract the preparation history task records of magnesium-based alloys from the company's internal production management system (such as ERP, MES, etc.), including task ID, preparation date, preparation batch, preparation process parameters, etc.

[0022] Raw material composition data: Obtain detailed composition data of magnesium-based alloy raw materials from the raw material supplier's database or the company's internal quality inspection system, such as the content of magnesium, aluminum, zinc, manganese, rare earth, titanium and other elements.

[0023] Performance test data: Obtain performance test reports of magnesium-based alloys from laboratories or quality inspection departments, including key performance indicators such as tensile strength, yield strength, elongation, and hardness.

[0024] Test condition data: record the environmental conditions during each performance test, such as test temperature, humidity, loading rate, etc.

[0025] Test equipment information: Get detailed information about the test equipment from the equipment management system, including equipment model, specifications, calibration records, etc.

[0026] Operator skill level data: Collect information such as operator training records, years of experience, and historical operation accuracy through the human resources management system or questionnaire survey.

[0027] Integrating the data from the above different sources to form a comprehensive data set containing complete information on the magnesium-based alloy preparation process requires the use of data integration tools or the writing of special data extraction, transformation, and loading (ETL) scripts.

[0028] Magnesium-based alloy dataset preprocessing: Check whether there are missing values ​​in the data set. Consider whether to remove data records with a large number of missing values. For a small number of missing values, interpolation methods (such as mean interpolation, median interpolation, etc.) can be used to fill them. Use statistical methods (such as box plots, Z-score, etc.) to detect outliers in the data set and correct or remove them according to actual conditions. Make the data of the same batch or the same task ID consistent in various indicators to avoid data entry errors.

[0029] For numerical features (such as element content, performance indicators, etc.), standardization processing (such as Min-Max scaling, Z-score standardization, etc.) is performed according to their value range to improve the accuracy and efficiency of subsequent data analysis. For categorical features (such as equipment model, operator experience, etc.), appropriate encoding conversion (such as one-hot encoding, label encoding, etc.) is performed to facilitate machine learning algorithm processing.

[0030] The preprocessed data set is divided into a certain proportion (such as 70% training set, 15% validation set, and 15% test set) to prepare for subsequent model training and verification.

[0031] Store the preprocessed data set in an appropriate data storage system (such as a relational database, NoSQL database, data warehouse, etc.) for subsequent calls and analysis. Establish a data dictionary and metadata management mechanism to improve data traceability and maintainability.

[0032] Step S102, extracting features from the preprocessed magnesium-based alloy data set to obtain magnesium-based alloy preparation history task features and magnesium-based alloy data features, the magnesium-based alloy data features include magnesium-based alloy data raw material composition features, magnesium-based alloy data performance test features, magnesium-based alloy data test condition features, magnesium-based alloy data test equipment features, and magnesium-based alloy data operator skill level features, and integrating the magnesium-based alloy data features to obtain magnesium-based alloy data statistical analysis results; Magnesium-based alloy preparation task ID: It is used to uniquely identify each magnesium-based alloy preparation task and is the basis for data traceability and management.

[0033] Preparation date of magnesium-based alloy: Recording the specific date of preparation of magnesium-based alloy will help analyze the impact of time factors on alloy quality.

[0034] Magnesium-based alloy production batches: Differentiate between different batches of magnesium-based alloys to facilitate tracking of quality differences between batches.

[0035] Magnesium-based alloy preparation process parameters: including melting temperature, casting speed, heat treatment conditions, etc. These parameters directly affect the final properties of the alloy.

[0036] Relevant information is directly extracted from the historical task database of magnesium-based alloy preparation. For process parameters, it is necessary to combine production records, process documents or on-site monitoring data for comprehensive extraction.

[0037] Content characteristics of each element in magnesium-based alloy data: including the content of major alloying elements such as magnesium, aluminum, zinc, manganese, rare earth, and titanium. The proportion of these elements directly affects the mechanical and physical properties of the alloy. Use chemical analysis methods (such as spectral analysis, energy spectrum analysis, etc.) to test raw materials and alloy samples to obtain the accurate content of each element. Digitize the test results and enter them into the magnesium-based alloy data set.

[0038] Magnesium-based alloy data performance test feature extraction: Magnesium-based alloy tensile strength characteristics: a measure of the alloy's ability to resist tensile fracture.

[0039] Yield strength characteristics of magnesium-based alloys: indicates the maximum stress of the alloy before plastic deformation occurs.

[0040] Elongation characteristics of magnesium-based alloys: reflects the plastic deformation ability of the alloy before fracture.

[0041] Hardness characteristics of magnesium-based alloys: Measures the ability of the alloy to resist local deformation, commonly expressed by indicators such as Brinell hardness and Rockwell hardness.

[0042] Relevant data is obtained through standard test methods such as tensile test and hardness test. The test process complies with relevant standards, the test equipment is calibrated, and the test results are accurate and reliable.

[0043] Magnesium-based alloy data test condition feature extraction: Temperature characteristics of magnesium-based alloy data testing: The ambient temperature during testing has a significant impact on the test results.

[0044] Humidity characteristics of magnesium-based alloy data test: humidity of the test environment, humidity of the test environment affects the corrosion performance of the alloy, etc.

[0045] Loading rate characteristics of magnesium-based alloy data tests: The loading speed in dynamic tests such as tensile tests affects the accuracy of the test results.

[0046] Parameters such as ambient temperature, humidity and loading rate were recorded during the test.

[0047] Magnesium-based alloy data test equipment feature extraction: Magnesium-based alloy data testing equipment model characteristics: Different models of testing equipment have different accuracy and stability.

[0048] Magnesium-based alloy data test specification characteristics: equipment specifications, such as test force range, displacement resolution, etc.

[0049] Magnesium-based alloy data test calibration record Features: Regular calibration records of equipment.

[0050] Extract equipment models, specifications, and calibration records from equipment archives or management systems.

[0051] Feature extraction of skill level of magnesium-based alloy data operators: Characteristics of training records for magnesium-based alloy data operators: training experience and skill level of operators.

[0052] Characteristics of years of experience of magnesium-based alloy data operators: years of experience engaged in magnesium-based alloy preparation and testing.

[0053] Historical operation accuracy characteristics of magnesium-based alloy data operators: error rate and accuracy rate in previous operations.

[0054] Extract operators’ training records and years of experience from the human resource management system. Evaluate operators’ accuracy by analyzing historical operation data.

[0055] Feature integration and statistical analysis results of magnesium-based alloy data: The extracted features are organized according to a certain data structure to form a magnesium-based alloy data set. The data set is cleaned and preprocessed to remove outliers and missing values.

[0056] Statistical analysis results: Statistical methods were used to analyze the magnesium-based alloy data set, and statistical quantities such as the mean, standard deviation, and correlation coefficient of each feature were calculated. The key factors and potential relationships that affect the quality of magnesium-based alloys were identified through data analysis. The statistical analysis results were used as the basis for subsequent neural network model training and quality rating.

[0057] Step S103, based on the magnesium-based alloy data characteristics and the magnesium-based alloy data statistical analysis results, a neural network model is used for training to establish a quality rating model for the magnesium-based alloy, where the quality rating model for the magnesium-based alloy is used to identify the relationship between raw material composition, test conditions and alloy quality; The statistical analysis results of the magnesium-based alloy data obtained in step S102 are used as the input of the neural network model. The statistical analysis results of the magnesium-based alloy data include characteristics of historical tasks in the preparation of magnesium-based alloys, characteristics of raw material composition, characteristics of performance tests, characteristics of test conditions, characteristics of test equipment, and characteristics of operator skill levels.

[0058] The input data is standardized to make the dimensions of different features consistent, to avoid some features having too much impact on model training due to too large or too small value ranges. Commonly used standardization methods include Z-score standardization or Min-Max standardization.

[0059] The standardized data is divided into training set, validation set and test set. Usually, the training set is used for model training, the validation set is used for model tuning, and the test set is used for final evaluation of model performance. The division ratio can be determined according to actual conditions, such as 70% training set, 15% validation set, and 15% test set.

[0060] Neural network model construction: Choose a neural network architecture: Choose a suitable neural network architecture based on the complexity of the problem and the size of the data. For the quality rating problem of magnesium-based alloys, you can consider using a multi-layer perceptron (MLP), convolutional neural network (CNN), or recurrent neural network (RNN). Since this problem mainly involves modeling nonlinear relationships between features, MLP is a more appropriate choice.

[0061] Determine the number of network layers and nodes: Determine the number of network layers and the number of nodes in each layer based on the number of input features and the complexity of the model. For example, a three-layer MLP structure can be used, where the number of input layer nodes is equal to the number of features, the number of hidden layer nodes can be determined based on experience or experiments, and the number of output layer nodes is the number of quality rating categories (such as excellent, qualified, unqualified, etc.).

[0062] Activation function selection: Use nonlinear activation functions (such as ReLU, Sigmoid, or Tanh) in the hidden layer to increase the expressiveness of the model. In the output layer, choose a suitable activation function (such as Softmax for multi-classification problems) based on the number of quality rating categories.

[0063] Select a suitable loss function (such as the cross entropy loss function) to measure the difference between the model prediction result and the true label. The optimization algorithm can choose stochastic gradient descent (SGD), Adam, etc. to minimize the loss function and update the model parameters.

[0064] Initialize model parameters: Randomly initialize the weights and bias parameters of the neural network.

[0065] Use the training set for forward propagation to calculate the loss between the model output and the true label. Use the backpropagation algorithm to calculate the gradient of the loss function with respect to the model parameters, and update the model parameters to reduce the loss.

[0066] Repeat the forward propagation and back propagation process until the preset training rounds are reached or the performance on the validation set no longer improves significantly. During the training process, early stopping can be used to prevent overfitting.

[0067] After each training cycle, use the validation set to evaluate model performance. Adjust the model architecture, hyperparameters (such as learning rate, batch size, etc.) or perform regularization (such as L2 regularization, Dropout, etc.) based on the performance on the validation set to prevent overfitting.

[0068] The trained neural network model is encapsulated as a quality rating model for magnesium-based alloys. The model can receive raw material composition data, performance test data, test condition data, test equipment information, and operator skill level data of magnesium-based alloys as input, and output the quality rating results of magnesium-based alloys.

[0069] The quality rating results of magnesium-based alloys can include quality scores, quality grades, and quality classification results, etc. These results will be used for subsequent magnesium-based alloy recommendations and preparation parameter optimization.

[0070] Step S104, obtaining historical data of magnesium-based alloy preparation, substituting the historical data of magnesium-based alloy preparation into a quality rating model of magnesium-based alloy, and obtaining a quality data set of magnesium-based alloy; Extract the historical data of magnesium-based alloy preparation from the preset database. The database should contain complete magnesium-based alloy preparation records, covering the entire process from raw material procurement to final product delivery.

[0071] The historical data of magnesium-based alloy preparation should include but not be limited to the following information: historical magnesium-based alloy raw material composition during the preparation process, historical magnesium-based alloy test conditions (such as temperature, humidity, loading rate, etc.), historical magnesium-based alloy test equipment information (such as equipment model, specifications, calibration records, etc.), historical magnesium-based alloy operator records (such as operator training records, years of experience, historical operation accuracy, etc.), and corresponding magnesium-based alloy performance test results (such as tensile strength, yield strength, elongation, hardness, etc.).

[0072] The extracted historical data of magnesium-based alloy preparation are cleaned to remove duplicate, erroneous or incomplete data records. The numerical data are normalized for subsequent model training. At the same time, the data are formatted according to the characteristic requirements of the magnesium-based alloy quality rating model.

[0073] According to the feature requirements of the magnesium-based alloy quality rating model, the required features are extracted from the pre-processed magnesium-based alloy preparation history data. These features include but are not limited to magnesium-based alloy raw material composition features, performance test features, test condition features, test equipment features, and operator skill level features.

[0074] The extracted features are integrated to form a complete feature set of historical data for magnesium-based alloy preparation. This step aims to convert the raw data into a format that the model can understand and process.

[0075] The integrated magnesium-based alloy preparation historical data feature set is used as input and substituted into the trained magnesium-based alloy quality rating model.

[0076] The magnesium-based alloy quality rating model calculates based on the input data characteristics through its internal neural network structure. The model uses the knowledge learned during the training process to classify and score the input data to obtain the quality rating results of the magnesium-based alloy.

[0077] The quality rating data output by the magnesium-based alloy quality rating model is sorted, including quality scores, quality grades, and quality classification results. The sorted quality rating data is associated with the original magnesium-based alloy preparation history data to form a quality data set for magnesium-based alloys. This step aims to establish a corresponding relationship between quality rating and preparation history, providing a basis for subsequent data analysis and recommendations.

[0078] Verify the generated magnesium-based alloy quality data set. This can be achieved by comparing it with known high-quality or low-quality magnesium-based alloy samples. Store the verified magnesium-based alloy quality data set in a database for subsequent data analysis and application. At the same time, update and maintain the database regularly.

[0079] Step S105, collecting raw material composition data in the magnesium-based alloy to be analyzed, matching the raw material composition data in the magnesium-based alloy to be analyzed with the quality data set of the magnesium-based alloy, obtaining the magnesium-based alloy quality level corresponding to the raw material composition data in the magnesium-based alloy to be analyzed, and using the magnesium-based alloy quality level corresponding to the raw material composition data in the magnesium-based alloy to be analyzed as the real-time magnesium-based alloy quality level; Select representative samples from the batch of magnesium-based alloy to be analyzed so that the samples can accurately reflect the overall characteristics of the batch of alloys.

[0080] Perform necessary pretreatment on samples, such as cutting, grinding, cleaning, etc., to improve the accuracy of component analysis.

[0081] Select appropriate raw material composition detection equipment, such as spectrometers, mass spectrometers, etc. These equipment should have high precision, high sensitivity and high stability.

[0082] The equipment should be calibrated before testing. The calibration process should follow the standards and procedures provided by the equipment manufacturer.

[0083] Use the selected component testing equipment to test the pretreated samples to obtain accurate content data of each element in the magnesium-based alloy (such as magnesium, aluminum, zinc, manganese, rare earth, titanium, etc.). Record various parameters during the testing process, such as test time, equipment status, environmental conditions, etc., for subsequent data tracing and analysis.

[0084] The raw data obtained from the test is processed to remove interference factors such as outliers and noise. The processed data is organized into formatted files or database records for subsequent data analysis and matching.

[0085] Match the raw material composition data of the magnesium-based alloy to be analyzed with the quality data set of the magnesium-based alloy: The raw material composition data of the magnesium-based alloy to be analyzed is preprocessed, including data cleaning, normalization, standardization and other operations, so as to effectively match the data in the quality data set of the magnesium-based alloy. Key features are extracted from the preprocessed raw material composition data of the magnesium-based alloy to be analyzed, and these features should be consistent with the features in the quality data set of the magnesium-based alloy, such as the content characteristics of each element.

[0086] Selection of similarity calculation algorithm: Select appropriate similarity calculation algorithms, such as Euclidean distance, cosine similarity, Pearson correlation coefficient, etc. These algorithms can accurately calculate the similarity between the data to be analyzed and the data in the quality data set. The extracted data characteristics of the raw material composition of the magnesium-based alloy to be analyzed are used as input, substituted into the selected similarity calculation algorithm, and matched one by one with the data in the magnesium-based alloy quality data set. According to the similarity calculation results, find the historical data records that match the data to be analyzed, and these records should contain the corresponding magnesium-based alloy quality level information.

[0087] The quality level information of the magnesium-based alloy is extracted from the historical data records matching the composition data of the raw materials of the magnesium-based alloy to be analyzed. The quality level is used as the real-time quality level of the magnesium-based alloy to be analyzed for subsequent preparation parameter recommendation and magnesium-based alloy selection.

[0088] Step S106, calling up the magnesium-based alloy preparation task corresponding to the real-time magnesium-based alloy quality level, taking the magnesium-based alloy preparation task corresponding to the real-time magnesium-based alloy quality level as the magnesium-based alloy preparation task to be simulated, substituting the magnesium-based alloy preparation task to be simulated into a preset magnesium-based alloy preparation simulation model, obtaining simulation parameters of the magnesium-based alloy preparation task to be simulated, comparing the simulation parameters of the magnesium-based alloy preparation task to be simulated with the real-time magnesium-based alloy quality level, and obtaining magnesium-based alloy preparation parameter error data; Construct a mapping table between quality levels and magnesium-based alloy preparation tasks. The mapping table between quality levels and magnesium-based alloy preparation tasks records the typical or historical magnesium-based alloy preparation tasks corresponding to each quality level (such as high quality, medium quality, low quality, etc.). These tasks should cover different alloy compositions, preparation processes, equipment conditions, etc., to fully reflect the relationship between quality levels and preparation parameters.

[0089] By analyzing historical data, the successful magnesium-based alloy preparation task records at each quality level are extracted, including preparation task ID, alloy composition, preparation process parameters, test conditions, etc., to form a mapping table and store it in the database.

[0090] In step S105, the real-time quality level of the magnesium-based alloy to be analyzed has been determined. Using this level, the corresponding magnesium-based alloy preparation task is queried in the constructed mapping table.

[0091] Through SQL query or NoSQL database query statement, all preparation task records matching the real-time quality level are retrieved in the mapping table to form a set of magnesium-based alloy preparation tasks to be simulated.

[0092] Prepare a preset magnesium-based alloy preparation simulation model. The model should be able to predict the performance parameters of magnesium-based alloys (such as tensile strength, yield strength, elongation, hardness, etc.) based on the input preparation task parameters (such as alloy composition, preparation process, test conditions, etc.).

[0093] Use physical models, numerical models or machine learning models (such as neural networks, support vector machines, etc.) to build simulation models for magnesium-based alloy preparation. The model should be fully trained to improve its prediction accuracy.

[0094] Each task parameter in the magnesium-based alloy preparation task set to be simulated is substituted into the preset simulation model in turn to perform simulation calculations.

[0095] For each preparation task in the set, the data such as alloy composition, preparation process parameters, and test conditions are extracted as input to the simulation model. Based on the input parameters, the model calculates the predicted performance parameters of the magnesium-based alloy through its internal algorithm to form a simulation parameter set.

[0096] Clarify the standard range of performance parameters for each quality level. These standard ranges should be determined based on historical data, industry standards or customer needs. For each quality level, define the minimum, maximum or allowable range of performance parameters such as tensile strength, yield strength, elongation, hardness, etc. to form a quality level standard table.

[0097] Each performance parameter in the simulation parameter set is compared with the standard range of the real-time magnesium-based alloy quality level to calculate the error value. For each performance parameter in the simulation parameter set (such as tensile strength, yield strength, etc.), the corresponding standard range is found from the quality level standard table. Then, the error (such as absolute error, relative error, etc.) between the simulation value and the standard range is calculated to form a magnesium-based alloy preparation parameter error data set.

[0098] Analyze the parameter error data set of magnesium-based alloy preparation, identify the parameters or tasks with large errors, and output the error analysis results. Statistically calculate the error distribution of each performance parameter, identify the parameters or tasks whose errors exceed the preset threshold. Output the error analysis report, including the error value, error source, and recommended improvement measures.

[0099] Step S107, matching the magnesium-based alloy preparation error data with a preset magnesium-based alloy preparation error range knowledge base to obtain a magnesium-based alloy preparation error result. If the magnesium-based alloy preparation error result exceeds the standard error value, then retrieving the data exceeding the standard error value in the magnesium-based alloy preparation error data, marking the data exceeding the standard error value as a first type of label, marking the data not exceeding the standard error value as a second type of label, and aggregating the data corresponding to the first type of label and the second type of label to obtain a dynamic label data set; Standardize each error index in the error data of magnesium-based alloy preparation. This may include converting the error values to a unified unit or range for comparison with the preset error range knowledge base.

[0100] The preset error range knowledge base should contain the standard error ranges for each error index of magnesium-based alloy preparation. These ranges can be determined based on historical data, industry standards, or expert experience. The standard error range for each error index can be expressed as an interval [L_i, U_i], where L_i is the lower limit and U_i is the upper limit.

[0101] For each error index in the error data of magnesium-based alloy preparation, compare it with the corresponding interval [L_i, U_i] in the preset error range knowledge base.

[0102] If the error value e_i satisfies L_i ≤ e_i ≤ U_i, it is considered that the error index is within the standard error range.

[0103] If the error value e_i does not satisfy the above conditions (i.e., e_i < L_i or e_i > U_i), it is considered that the error index exceeds the standard error range.

[0104] According to the matching results, classify each error index in the error data of magnesium-based alloy preparation into two categories: The first type of label: The error index data items that exceed the standard error value. These data items indicate significant deviations in the magnesium-based alloy preparation process and require special attention.

[0105] The second type of label: The error index data items that do not exceed the standard error value. These data items indicate that the magnesium-based alloy preparation process is relatively stable and meets the expected standards.

[0106] Summarize all the error index data items marked with the first type of label and the second type of label to form a dynamic label data set. This set contains the classification information of all error indexes in the current magnesium-based alloy preparation process and provides an important basis for subsequent optimization and recommendation.

[0107] Record the summary result of the dynamic label data set and feedback it to the preparation personnel or the quality control department so that they can understand the error situation in the current preparation process and take corresponding measures for improvement.

[0108] Step S108, substitute the dynamic tag data set into the preset magnesium-based alloy preparation model, output magnesium-based alloy preparation simulation data, optimize the magnesium-based alloy preparation simulation data based on the use of the predator optimization algorithm of the preset magnesium-based alloy preparation model to obtain optimized magnesium-based alloy preparation simulation data, and perform data processing on the optimized magnesium-based alloy preparation simulation data through the preset magnesium-based alloy production model to obtain optimized magnesium-based alloy preparation simulation parameters, match the optimized magnesium-based alloy preparation simulation parameters with the data corresponding to the real-time magnesium-based alloy quality level, obtain magnesium-based alloy preparation parameter data matching the real-time magnesium-based alloy quality level, and use the magnesium-based alloy preparation parameter data matching the real-time magnesium-based alloy quality level as recommended parameters; The dynamic label data set generated in step S107 is used as input and substituted into a preset magnesium-based alloy preparation model. The magnesium-based alloy preparation model is a mathematical model or simulation model constructed based on historical data and professional knowledge, which can simulate the preparation process of magnesium-based alloys and output corresponding simulation data.

[0109] Through the magnesium-based alloy preparation model, a set of simulation data can be obtained, which reflects the preparation results of magnesium-based alloy under the given dynamic label data set conditions. The simulation data includes the composition of the alloy, preparation process parameters, expected performance indicators, etc.

[0110] Next, the simulated data is optimized using the predator-prey optimization algorithm, which is a heuristic optimization algorithm that can simulate the behavior of predators in nature and find the optimal solution through continuous iteration and search.

[0111] During the optimization process, the simulation data is used as the input of the algorithm, and appropriate objective functions and constraints are set. Through the operation of the algorithm, the optimized magnesium-based alloy preparation simulation data is obtained.

[0112] The optimized magnesium-based alloy preparation simulation data is substituted into the preset magnesium-based alloy production model for data processing.

[0113] The magnesium-based alloy production model is a model used to process and analyze simulation data. It can extract key production parameters from the optimized simulation data.

[0114] Through data processing, optimized simulation parameters for magnesium-based alloy preparation can be obtained, which are the basis for subsequent matching and recommendation.

[0115] The optimized magnesium-based alloy preparation simulation parameters are matched with the data corresponding to the real-time magnesium-based alloy quality level.

[0116] The matching process can use similarity calculation, distance measurement and other methods to find the real-time magnesium-based alloy quality level data that matches the optimized simulation parameters.

[0117] Through the matching process, magnesium-based alloy preparation parameter data that matches the real-time magnesium-based alloy quality level can be obtained. The magnesium-based alloy preparation parameter data that matches the real-time magnesium-based alloy quality level is optimized and verified, and can enable the magnesium-based alloy preparation process to achieve the expected quality level. Therefore, these parameter data are used as recommended parameters for reference and use by preparation personnel or quality control departments.

[0118] Step S109, matching the recommended parameters with the historical data of magnesium-based alloy preparation, and using the magnesium-based alloy prepared by using the recommended parameters as the recommended magnesium-based alloy.

[0119] In step S109, the goal is to perform similarity matching between the recommended parameters and the historical data of magnesium-based alloy preparation, find preparation data matching the recommended parameters, and use the magnesium-based alloy prepared based on the recommended parameters as the recommended magnesium-based alloy.

[0120] Collect and organize the historical data of magnesium-based alloy preparation, which includes preparation records under different preparation conditions (such as alloy composition, preparation process, equipment parameters, etc.), as well as the corresponding magnesium-based alloy performance data (such as tensile strength, yield strength, elongation, etc.). Use the similarity calculation algorithm to compare the recommended parameters with each record in the historical data of magnesium-based alloy preparation. Similarity calculation can use cosine similarity, Euclidean distance, Manhattan distance and other methods, and select the appropriate algorithm according to the characteristics and requirements of the data. Through similarity calculation, find the historical preparation data records that match the recommended parameters.

[0121] According to the similarity calculation results, several historical preparation data records with the highest similarity are selected as recommended parameters to match the magnesium-based alloy preparation data.

[0122] Based on the selected recommended parameters, the magnesium-based alloy preparation data is matched and the actual magnesium-based alloy preparation process parameters are adjusted to ensure that the alloy composition, preparation process, equipment parameters, etc. in the preparation process are consistent or similar to the matching data.

[0123] The adjusted magnesium-based alloy preparation process is implemented to prepare magnesium-based alloy samples. The prepared magnesium-based alloy samples are subjected to performance testing and evaluation to verify whether they meet the expected quality level and performance requirements.

[0124] The performance data of the prepared magnesium-based alloy samples were compared with the recommended parameters and matching historical data to verify the effectiveness of similarity matching and the accuracy of the preparation process.

[0125] According to the verification results, the recommendation parameters and similarity matching algorithm are adjusted and optimized to improve the accuracy and preparation efficiency of subsequent recommendations.

[0126] Specifically, the quality rating and recommendation method of magnesium-based alloys according to the present invention, step S102, comprises: The characteristics of the magnesium-based alloy preparation history task include the magnesium-based alloy preparation task ID, magnesium-based alloy preparation date, magnesium-based alloy preparation batch, and magnesium-based alloy preparation process parameters; The raw material composition characteristics of magnesium-based alloy data include the content characteristics of each element of magnesium-based alloy data; The data performance test characteristics of magnesium-based alloys include magnesium-based alloy tensile strength characteristics, magnesium-based alloy yield strength characteristics, magnesium-based alloy elongation characteristics and magnesium-based alloy hardness characteristics; The test condition characteristics of magnesium-based alloy data include the test temperature characteristics of magnesium-based alloy data, the test humidity characteristics of magnesium-based alloy data, and the test loading rate characteristics of magnesium-based alloy data; The characteristics of the magnesium-based alloy data testing equipment include the characteristics of the magnesium-based alloy data testing equipment model, the characteristics of the magnesium-based alloy data testing specification, and the characteristics of the magnesium-based alloy data testing calibration record; The skill level characteristics of magnesium-based alloy data operators include the training record characteristics of magnesium-based alloy data operators, the years of experience characteristics of magnesium-based alloy data operators, and the historical operation accuracy characteristics of magnesium-based alloy data operators; The statistical analysis results of magnesium-based alloy data are obtained by integrating the characteristics of historical tasks in magnesium-based alloy preparation, raw material composition characteristics of magnesium-based alloy data, performance test characteristics of magnesium-based alloy data, test condition characteristics of magnesium-based alloy data, test equipment characteristics of magnesium-based alloy data and skill level characteristics of magnesium-based alloy data operators.

[0127] In step S102, the magnesium-based alloy quality rating and recommendation method of the present invention involves integrating various features in the magnesium-based alloy preparation and testing process to obtain statistical analysis results of magnesium-based alloy data.

[0128] Integration of historical mission characteristics of magnesium-based alloy preparation: Magnesium-based alloy preparation task ID: Assign a unique identifier to each magnesium-based alloy preparation task to facilitate subsequent data tracking and management.

[0129] Preparation date of magnesium-based alloy: record the preparation date of magnesium-based alloy in order to analyze the influence of preparation time on alloy properties.

[0130] Magnesium-based alloy preparation batch: Identify magnesium-based alloys prepared in the same batch to improve the traceability and consistency of the same batch of alloys.

[0131] Magnesium-based alloy preparation process parameters: Detailed records of key process parameters in the preparation process, such as temperature, pressure, time, etc., which have a direct impact on the performance of the alloy.

[0132] Magnesium-based alloy data raw material composition characteristics integration: Content characteristics of each element in magnesium-based alloy preparation: Accurately determine the content of each element in magnesium-based alloy, including major elements and trace elements, through chemical analysis or spectral analysis and other methods.

[0133] Magnesium-based alloy data performance test feature integration: Tensile strength characteristics: The tensile strength of magnesium-based alloys is measured through tensile tests, which reflects the alloy's resistance to tension.

[0134] Yield strength characteristics: Determine the stress value of the alloy at the yield point, reflecting the resistance of the alloy before plastic deformation.

[0135] Elongation characteristic: Measures the degree of plastic deformation of the alloy before fracture, reflecting the plastic properties of the alloy.

[0136] Hardness characteristics: The hardness value of the alloy is measured through a hardness test, which reflects the alloy's ability to resist local deformation.

[0137] Magnesium-based alloy data test condition feature integration: Test temperature characteristics: record the ambient temperature during the test, because temperature has an impact on the performance test results of the alloy.

[0138] Test humidity characteristics: record the ambient humidity during the test. Humidity affects the surface state of the alloy and the test results.

[0139] Test loading rate characteristics: Record the loading rate during the test, as the loading rate will affect the mechanical properties test results of the alloy.

[0140] Magnesium-based alloy data testing equipment feature integration: Test equipment model characteristics: Record the model of the test equipment used.

[0141] Test Specifications: Describe the test equipment specifications and parameters in detail so that the test results can be reproduced.

[0142] Test calibration record feature: Record the calibration history and calibration results of the test equipment.

[0143] Integration of skill level characteristics of magnesium-based alloy data operators; Training record features: record the operator’s training experience and training content.

[0144] Years of experience feature: records the operator’s years of experience. The longer the years of experience, the higher the operator’s skill level.

[0145] Historical operation accuracy feature: It counts the accuracy of operators’ past operation tasks as an indicator to evaluate their skill level.

[0146] Data integration and statistical analysis: All the above characteristic data are integrated into one data set. The integrated data are analyzed in depth using statistical analysis methods, such as descriptive statistics, correlation analysis, regression analysis, etc. Based on the analysis results, useful information for quality rating and recommendation of magnesium-based alloys is extracted to form the statistical analysis results of magnesium-based alloy data.

[0147] Specifically, the quality rating and recommendation method of magnesium-based alloys according to the present invention, step S103, comprises: Using the statistical analysis results of magnesium-based alloy data as the input of the neural network model, initializing the parameters of the neural network model, and obtaining the initialized neural network model; The statistical analysis results of magnesium-based alloy data are divided into a training set and a validation set. The initialized neural network model is trained using the training set. The output is calculated through forward propagation, the error is calculated according to the loss function, and the model parameters are updated through back propagation. The training process is continuously iterated until the preset training rounds are reached or the performance on the validation set is no longer improved, then the trained neural network model is output; The trained neural network model is encapsulated as a quality rating model for magnesium-based alloys. The raw material composition data of magnesium-based alloys, performance test data of magnesium-based alloys, test condition data of magnesium-based alloys, test equipment information of magnesium-based alloys, and skill level data of magnesium-based alloy preparation operators are used as input, and the rating results of the quality of magnesium-based alloys are used as output.

[0148] Initialization of the neural network model: Input data preparation: The statistical analysis results of magnesium-based alloy data are used as the input data of the neural network model. These data include magnesium-based alloy raw material composition, performance test results, test conditions, test equipment information, and the skill level of the preparation operator.

[0149] Model structure design: Design the structure of the neural network model based on the characteristics of the input data and the requirements of quality rating. This includes determining the number of layers in the network, the number of neurons in each layer, the selection of activation functions, etc.

[0150] Parameter initialization: Initialize the parameters of the neural network model. This includes setting the initial values ​​of weights and biases, usually using random initialization or initialization based on a certain distribution (such as Gaussian distribution).

[0151] Get the initialized neural network model: After completing the parameter initialization, get an initialized neural network model, ready for training.

[0152] Model training process: Data segmentation: The statistical analysis results of magnesium-based alloy data are divided into a training set and a validation set. The training set is used to train the neural network model, and the validation set is used to evaluate the performance of the model.

[0153] Training starts: Use the training set to train the initialized neural network model.

[0154] Forward propagation calculation output: Input the data in the training set into the neural network model, and obtain the output of the model through forward propagation calculation.

[0155] Loss function calculation error: According to the output of the model and the true label (i.e., the quality rating result of magnesium-based alloy), the value of the loss function is calculated as the error of the model.

[0156] Back propagation updates model parameters: According to the gradient of the loss function, the parameters of the neural network model, including weights and biases, are updated through the back propagation algorithm.

[0157] Iterative training: Repeat the process of forward propagation, loss function calculation and back propagation to continuously iterate and train the neural network model.

[0158] Training termination condition: Set the performance improvement threshold on the training round or validation set as the training termination condition. When the preset training round or the performance on the validation set no longer improves, training stops.

[0159] Model output: Outputting a trained neural network model: After training, a trained neural network model is obtained, which can predict the quality rating results of magnesium-based alloys based on input data.

[0160] Model packaging: The trained neural network model is packaged into a quality rating model for magnesium-based alloys. The packaging process includes packaging the model's structure, parameters, and calculation logic into a callable module or interface.

[0161] Model usage: The encapsulated magnesium-based alloy quality rating model can receive magnesium-based alloy raw material composition data, performance test data, test condition data, test equipment information and preparation operator skill level data as input, and output the rating results of magnesium-based alloy quality.

[0162] Specifically, the quality rating and recommendation method of magnesium-based alloys according to the present invention, step S104, comprises: Obtaining historical preparation data of magnesium-based alloys from a preset database, the historical preparation data of magnesium-based alloys including historical magnesium-based alloy raw material composition, historical magnesium-based alloy test conditions, historical magnesium-based alloy test equipment information, and historical magnesium-based alloy operator records during the preparation process; Preprocessing the preparation history data of the magnesium-based alloy to obtain the preprocessed preparation history data of the magnesium-based alloy; According to the feature requirements of the magnesium-based alloy quality rating model, the features required by the magnesium-based alloy quality rating model are extracted from the pre-processed magnesium-based alloy preparation history data to obtain an extracted magnesium-based alloy data feature set. The extracted magnesium-based alloy data feature set includes the features required by the magnesium-based alloy quality rating model, including magnesium-based alloy raw material composition data features, magnesium-based alloy performance test data features, magnesium-based alloy test condition data features, magnesium-based alloy test equipment information features, and magnesium-based alloy preparation operator skill level data features; The extracted magnesium-based alloy data feature set is input into the magnesium-based alloy quality rating model. The magnesium-based alloy quality rating model calculates the quality rating of the magnesium-based alloy through the neural network structure inside the magnesium-based alloy quality rating model according to the input data. The quality rating data of the magnesium-based alloy includes the quality score of the magnesium-based alloy, the quality grade of the magnesium-based alloy and the quality classification result of the magnesium-based alloy; The quality rating data of the magnesium-based alloy output by the magnesium-based alloy quality rating model is sorted out, and the quality data of the magnesium-based alloy is associated with the original magnesium-based alloy preparation history data to obtain a quality data set of the magnesium-based alloy.

[0163] In step S104, the magnesium-based alloy quality rating and recommendation method of the present invention involves obtaining the preparation history data of the magnesium-based alloy from a database, and finally obtaining a quality data set of the magnesium-based alloy through a series of processing steps.

[0164] Get historical data on magnesium-based alloy preparation: Database connection and query: First, establish a connection with the preset database and perform a query operation to obtain the preparation history data of magnesium-based alloys.

[0165] Data content: The acquired data should include historical magnesium-based alloy raw material composition during the preparation process, historical magnesium-based alloy test conditions, historical magnesium-based alloy test equipment information, and historical magnesium-based alloy operator records.

[0166] Preprocessing of historical data of magnesium-based alloy preparation: Data cleaning: Clean the acquired data to remove duplicate, missing or abnormal data records.

[0167] Data formatting: Convert data into a unified format to facilitate subsequent processing and analysis.

[0168] Data normalization / standardization: Normalize or standardize numerical data to eliminate the impact of dimensional differences on the model.

[0169] Obtaining the historical data of magnesium-based alloy preparation after pretreatment: After the above processing steps, the historical data of magnesium-based alloy preparation after pretreatment is obtained.

[0170] Feature requirement analysis: According to the feature requirements of the magnesium-based alloy quality rating model, the feature types that need to be extracted are clarified.

[0171] Feature extraction operation: Extract the features required for the magnesium-based alloy quality rating model from the preprocessed magnesium-based alloy preparation history data, including magnesium-based alloy raw material composition data features, magnesium-based alloy performance test data features, magnesium-based alloy test condition data features, magnesium-based alloy test equipment information features, and magnesium-based alloy preparation operator skill level data features.

[0172] Obtain the extracted magnesium-based alloy data feature set: Integrate the extracted features into a feature set as the input of the magnesium-based alloy quality rating model.

[0173] Calculation of magnesium-based alloy quality rating model: Model input: The extracted magnesium-based alloy data feature set is input into the magnesium-based alloy quality rating model.

[0174] Model calculation: The magnesium-based alloy quality rating model performs calculations based on the input data through its internal neural network structure.

[0175] Obtain quality rating: After the calculation is completed, the model outputs the quality rating data of the magnesium-based alloy, including the quality score, quality grade and quality classification results of the magnesium-based alloy.

[0176] Quality data collation and association: Data collation: collate the quality rating data output by the magnesium-based alloy quality rating model.

[0177] Data association: Associate the sorted magnesium-based alloy quality data with the original magnesium-based alloy preparation history data to establish a corresponding relationship between the two.

[0178] Obtaining a quality data set of the magnesium-based alloy: After the above processing steps, a quality data set containing quality rating information of the magnesium-based alloy is obtained.

[0179] Specifically, the quality rating and recommendation method of magnesium-based alloys according to the present invention, step S105, comprises: Use magnesium-based alloy raw material composition testing equipment to test the magnesium-based alloy sample to be analyzed, and collect raw material composition data of the magnesium-based alloy to be analyzed; Preprocessing the raw material composition data of the magnesium-based alloy to be analyzed, matching the preprocessed raw material composition data of the magnesium-based alloy to be analyzed with the data in the magnesium-based alloy quality data set, and obtaining historical data records matching the raw material composition of the magnesium-based alloy to be analyzed based on a preset similarity calculation algorithm; Extracting historical data of quality level of magnesium-based alloy from historical data records matching the raw material composition of magnesium-based alloy to be analyzed, and obtaining the quality level of magnesium-based alloy corresponding to the raw material composition data in magnesium-based alloy to be analyzed; The magnesium-based alloy quality level corresponding to the raw material composition data in the magnesium-based alloy to be analyzed is used as the real-time magnesium-based alloy quality level of the magnesium-based alloy to be analyzed.

[0180] In step S105, the magnesium-based alloy quality rating and recommendation method of the present invention involves testing the raw material composition of the magnesium-based alloy sample to be analyzed, and determining the real-time quality level of the magnesium-based alloy to be analyzed by matching the raw material composition with the data in the magnesium-based alloy quality data set.

[0181] Composition test of raw materials of magnesium-based alloy to be analyzed: Equipment preparation: Keep the magnesium-based alloy raw material composition testing equipment in good working condition and calibrate the equipment to improve detection accuracy.

[0182] Sample testing: Place the magnesium-based alloy sample to be analyzed into the testing equipment and perform the test according to the equipment operating procedures.

[0183] Data collection: The equipment automatically collects the raw material composition data of the magnesium-based alloy to be analyzed, including the content of various elements.

[0184] Raw material composition data preprocessing: Data cleaning: Clean the collected raw material composition data to remove outliers or noise data.

[0185] Data formatting: Convert the data into a format consistent with the data in the magnesium-based alloy quality data set to facilitate subsequent matching operations.

[0186] Data normalization / standardization (if necessary): Normalize or standardize the data as needed to eliminate the impact of dimension differences on similarity calculation.

[0187] Data matching similarity calculation: Data matching preparation: The pre-processed raw material composition data of the magnesium-based alloy to be analyzed is compared with the data in the magnesium-based alloy quality data set.

[0188] Similarity calculation algorithm selection: Select an appropriate similarity calculation algorithm according to actual needs, such as cosine similarity, Euclidean distance, etc.

[0189] Similarity calculation: Use the selected similarity calculation algorithm to calculate the similarity between the raw material composition data of the magnesium-based alloy to be analyzed and each historical data record in the magnesium-based alloy quality data set.

[0190] Matching historical data records are determined: based on the similarity calculation result, historical data records that match the composition of the raw material of the magnesium-based alloy to be analyzed are determined, and the matched historical data are historical data records that match the composition of the raw material of the magnesium-based alloy to be analyzed.

[0191] Magnesium-based alloy quality grade extraction: Quality level historical data extraction: Extract the quality level historical data of magnesium-based alloy from the historical data records that match the composition of the raw materials of the magnesium-based alloy to be analyzed.

[0192] Quality level determination: Determine the quality level of the magnesium-based alloy corresponding to the raw material composition data in the magnesium-based alloy to be analyzed based on the extracted quality level historical data.

[0193] Real-time magnesium-based alloy quality grade determination: Quality level assignment: The determined magnesium-based alloy quality level is used as the real-time magnesium-based alloy quality level of the magnesium-based alloy to be analyzed.

[0194] Result output: The real-time magnesium-based alloy quality level is output to relevant users or systems for subsequent production decisions, quality control or product recommendations, etc.

[0195] Specifically, the quality rating and recommendation method of magnesium-based alloys according to the present invention, step S106, comprises: According to the real-time magnesium-based alloy quality level, query and retrieve the magnesium-based alloy preparation task record corresponding to the real-time magnesium-based alloy quality level in a preset database to obtain the magnesium-based alloy preparation task corresponding to the real-time magnesium-based alloy quality level; The magnesium-based alloy preparation task corresponding to the real-time magnesium-based alloy quality level is determined as the magnesium-based alloy preparation task to be simulated, and the magnesium-based alloy preparation task to be simulated is substituted into a preset magnesium-based alloy preparation simulation model, and the preset magnesium-based alloy preparation simulation model outputs simulation parameters of the magnesium-based alloy preparation task to be simulated, and the simulation parameters of the magnesium-based alloy preparation task to be simulated include the expected alloy properties of magnesium-based alloy preparation, variables in the magnesium-based alloy preparation process, control parameters of magnesium-based alloy preparation equipment, and content parameters of each component in magnesium-based alloy preparation; The simulation parameters of the magnesium-based alloy preparation task to be simulated are compared with the real-time magnesium-based alloy quality level to obtain the comparison result of the simulation parameters of the magnesium-based alloy preparation task to be simulated and the real-time magnesium-based alloy quality level, that is, the magnesium-based alloy preparation parameter error data. The magnesium-based alloy preparation parameter error data includes the magnesium-based alloy preparation error value, the magnesium-based alloy preparation error range and the magnesium-based alloy preparation error source analysis.

[0196] In step S106, the magnesium-based alloy quality rating and recommendation method described in the present invention involves querying and retrieving corresponding magnesium-based alloy preparation task records according to the real-time magnesium-based alloy quality level, and simulating through a magnesium-based alloy preparation simulation model to obtain preparation parameter error data.

[0197] Query and retrieve magnesium-based alloy preparation task records: Database connection: Establish a connection with the preset database to enable access to the database that stores magnesium-based alloy preparation task records.

[0198] Query condition setting: According to the real-time magnesium-based alloy quality level, set the query conditions, such as quality level code, quality score range, etc.

[0199] Task record query: perform a query operation to retrieve the magnesium-based alloy preparation task records corresponding to the real-time magnesium-based alloy quality level from the database.

[0200] Task record retrieval: retrieve the queried magnesium-based alloy preparation task records as input for subsequent simulations.

[0201] Determine the magnesium-based alloy preparation task to be simulated: Task filtering: If the query results contain multiple preparation task records, filter out the preparation task that best suits the current situation as the task to be simulated based on actual needs or preset rules.

[0202] Task determination: The screened magnesium-based alloy preparation task is determined as the magnesium-based alloy preparation task to be simulated.

[0203] Substitute into the simulation model for magnesium-based alloy preparation: Model preparation: Ensure that the preset magnesium-based alloy preparation simulation model has been correctly configured and can receive the task to be simulated as input.

[0204] Task input: Substitute the magnesium-based alloy preparation task to be simulated into the magnesium-based alloy preparation simulation model.

[0205] Simulation execution: Start the simulation model, execute the simulation process, calculate and output the simulation parameters of the magnesium-based alloy preparation task to be simulated.

[0206] Output simulation parameters: Parameter type determination: Clarify the type of simulation parameters, including the expected alloy properties of magnesium-based alloy preparation, variables in the preparation process, control parameters of the preparation equipment, and parameters of the content of each component.

[0207] Parameter output: After the simulation model calculation is completed, the simulation parameters of the magnesium-based alloy preparation task to be simulated are output.

[0208] Comparison of simulation parameters with real-time magnesium alloy quality levels: Comparison preparation: Preparation for comparing simulation parameters with real-time magnesium-based alloy quality levels to make the two comparable.

[0209] Compare Execute: Perform a compare operation to calculate the difference between the simulation parameters and the real-time magnesium-based alloy quality level.

[0210] Comparison result output: obtain the comparison result between the simulation parameters of the magnesium-based alloy preparation task to be simulated and the real-time magnesium-based alloy quality level, that is, the magnesium-based alloy preparation parameter error data.

[0211] Analysis of magnesium-based alloy preparation parameter error data: Error value calculation: Calculate the error value in the magnesium-based alloy preparation process to reflect the degree of difference between the simulation parameters and the real-time quality level.

[0212] Error range determination: Determine the range of error values, such as maximum error, minimum error and average error.

[0213] Error source analysis: Conduct in-depth analysis of the error sources to identify the main causes of the errors, such as fluctuations in raw material composition, inaccurate preparation process control, equipment failure, etc.

[0214] Specifically, the quality rating and recommendation method of magnesium-based alloys according to the present invention, step S107, comprises: Matching each error index in the magnesium-based alloy preparation error data with the standard error range in the preset error range knowledge base one by one, and obtaining a matching result, i.e., a magnesium-based alloy preparation error result, wherein the magnesium-based alloy preparation error result includes each magnesium-based alloy preparation error index; Determine whether there are indicators exceeding the standard error value for each magnesium-based alloy preparation error index in the magnesium-based alloy preparation error result, and obtain magnesium-based alloy preparation error index data items exceeding the standard error value; Retrieving magnesium-based alloy preparation error index data items exceeding the standard error value from magnesium-based alloy preparation error data; The magnesium-based alloy preparation error index data items exceeding the standard error value are marked as first-class labels, and the magnesium-based alloy preparation error index data items not exceeding the standard error value are marked as second-class labels; The first category label data and the second category label data are aggregated to form a data set including the first category label data and the second category label data, and the data set including the first category label data and the second category label data is used as a dynamic label data set.

[0215] In step S107, the magnesium-based alloy quality rating and recommendation method described in the present invention involves matching the magnesium-based alloy preparation error data with a preset error range knowledge base, and classifying and marking the error indicators according to the matching results, and finally forming a dynamic label data set.

[0216] Matching magnesium-based alloy preparation error data and standard error range: Error index extraction: Extract various error indicators from the error data of magnesium-based alloy preparation, such as error value, error range, etc.

[0217] Knowledge base connection: Connect to the preset error range knowledge base to access the standard error range data therein.

[0218] One-to-one matching: Match the extracted error indicators with the standard error range in the knowledge base one by one, and compare the actual error value with the standard error value.

[0219] Matching result output: The matching result, i.e., the magnesium-based alloy preparation error result, is obtained, which includes the comparison between various magnesium-based alloy preparation error indicators and the standard error range.

[0220] Determine whether the error index exceeds the standard error value: Judgment standard setting: clearly define the standard for judging whether the error indicator exceeds the standard error value. If the error value is greater than the standard error value, it is considered to exceed the standard error value.

[0221] Item-by-item judgment: Each error indicator in the error results of magnesium-based alloy preparation is judged item by item to determine whether there is an indicator that exceeds the standard error value.

[0222] Extraction of out-of-index data items: Extract the magnesium-based alloy preparation error index data items that exceed the standard error value.

[0223] Retrieve and mark the error indicator data items: Retrieve data items: retrieve error index data items that exceed the standard error value from the magnesium-based alloy preparation error data.

[0224] Marking labels: Error indicator data items that exceed the standard error value are marked as the first type of labels, indicating that the error items need to be paid special attention to or adjusted; error indicator data items that do not exceed the standard error value are marked as the second type of labels, indicating that the error items are within the normal range.

[0225] Aggregate the labeled data to form a data set: Data aggregation: Aggregate the first-category labeled data and the second-category labeled data so that all data items are included.

[0226] Data set formation: A data set including first-category label data and second-category label data is formed, and the data set can fully reflect the error situation in the magnesium-based alloy preparation process.

[0227] Naming of dynamic label data set: The formed data set is named as dynamic label data set, indicating that the data set is dynamically updated as the preparation process proceeds.

[0228] Application of dynamic label data collection: Data analysis: Analyze the dynamic label data set to understand the error distribution and main error sources in the magnesium-based alloy preparation process.

[0229] Decision support: Based on the analysis results of the dynamic label data set, provide decision support for the optimization of the preparation process, the formulation of quality control measures, etc.

[0230] Continuous improvement: Continuous improvement and optimization of the magnesium-based alloy preparation process can be achieved by continuously monitoring and updating the dynamic tag data set.

[0231] Specifically, the quality rating and recommendation method of magnesium-based alloys according to the present invention, step S108, comprises: Substituting the data in the dynamic label data set into a preset magnesium-based alloy preparation model as input parameters of the magnesium-based alloy preparation model, the preset magnesium-based alloy preparation model preliminarily simulates the preparation results according to the input parameters to obtain magnesium-based alloy preparation simulation data; Apply the predatory optimization algorithm to optimize the simulation data of magnesium-based alloy preparation, and output the optimized simulation data of magnesium-based alloy preparation; Substituting the optimized magnesium-based alloy preparation simulation data into a preset magnesium-based alloy production model, and outputting the optimized magnesium-based alloy preparation simulation parameters; Matching the optimized magnesium-based alloy preparation simulation parameters with the data corresponding to the real-time magnesium-based alloy quality level to obtain magnesium-based alloy preparation parameter data matching the real-time magnesium-based alloy quality level; The magnesium-based alloy preparation parameter data matching the real-time magnesium-based alloy quality level is determined as the recommended parameters, and the recommended parameters are the process parameters for the preparation of the magnesium-based alloy.

[0232] In step S108, the magnesium-based alloy quality rating and recommendation method described in the present invention involves substituting the dynamic label data set into the magnesium-based alloy preparation model for simulation, optimizing the simulation data using a predatory optimization algorithm, and substituting the optimized results into the production model to obtain recommended parameters.

[0233] Substitute the dynamic label data set into the magnesium-based alloy preparation model: Data preparation: Extract the required data from the dynamically labeled data set.

[0234] Model connection: Connect the preset magnesium-based alloy preparation model so that the model can receive dynamically labeled datasets as input.

[0235] Data substitution: Substitute the extracted data into the magnesium-based alloy preparation model as the input parameters of the model.

[0236] Preliminary simulation results of magnesium-based alloy preparation: Simulation execution: Start the magnesium-based alloy preparation model and perform a preliminary simulation based on the input parameters.

[0237] Simulation data output: After the model calculation is completed, the simulation data of magnesium-based alloy preparation is output, including the expected alloy properties, variables in the preparation process, etc.

[0238] Apply the predator optimization algorithm to optimize the simulated data: Algorithm selection: Select a suitable predator optimization algorithm, such as genetic algorithm, particle swarm optimization algorithm, etc., to optimize the simulation data.

[0239] Optimization execution: Apply the selected predation optimization algorithm to optimize the simulation data of magnesium-based alloy preparation and find a better solution.

[0240] Optimized data output: After the algorithm runs, the optimized magnesium-based alloy preparation simulation data is output.

[0241] Substitute the optimized simulation data into the magnesium-based alloy production model; Fabrication Model Connection: Connect the preset magnesium-based alloy fabrication model so that the model can receive optimized simulation data as input.

[0242] Data substitution: Substitute the optimized magnesium-based alloy preparation simulation data into the production model.

[0243] Output optimized magnesium-based alloy preparation simulation parameters: Simulation execution: Start the magnesium-based alloy production model and perform simulation based on the input optimization data.

[0244] Parameter output: After the model calculation is completed, the optimized magnesium-based alloy preparation simulation parameters are output, including specific process parameters, equipment control parameters, etc.

[0245] Matching optimized simulation parameters with real-time magnesium-based alloy quality grade data: Data preparation: Extract data corresponding to the real-time quality level of magnesium-based alloys.

[0246] Matching execution: Match the optimized magnesium-based alloy preparation simulation parameters with the real-time magnesium-based alloy quality grade data to find matching parameter combinations.

[0247] Matching result output: obtain magnesium-based alloy preparation parameter data that matches the real-time magnesium-based alloy quality level.

[0248] Determine the recommended parameters: Recommended parameter selection: According to the matching results, select the matching magnesium-based alloy preparation parameter data as the recommended parameters.

[0249] Recommended parameter output: Output recommended parameters to guide the preparation process of magnesium-based alloys so that the quality of the prepared magnesium-based alloys meets expectations.

[0250] Specifically, the quality rating and recommendation method of magnesium-based alloys according to the present invention, step S109, comprises: According to the magnesium-based alloy preparation data, a similarity calculation model is constructed, and the characteristic vector of each data is extracted from the magnesium-based alloy preparation history data to obtain the characteristic vector of magnesium-based alloy preparation, and the characteristic vector of each data is extracted from the recommended parameters to obtain the characteristic vector of the recommended parameters; Using the constructed similarity calculation model, the feature vector of the recommended parameters is compared with the feature vector of the magnesium-based alloy preparation, and the similarity score between the feature vector of the recommended parameters and the feature vector of the magnesium-based alloy preparation is calculated; According to the similarity score between the characteristic vector of the recommended parameter and the characteristic vector of the magnesium-based alloy preparation, the historical data of the magnesium-based alloy preparation are sorted to obtain the sorting result of the historical data of the magnesium-based alloy preparation, and the data ranked first in the sorting result of the historical data of the magnesium-based alloy preparation is used as the data to be processed; The magnesium-based alloy finished product information corresponding to the data to be processed is retrieved from the magnesium-based alloy preparation history data, and the magnesium-based alloy finished product information corresponding to the data to be processed is used as the recommended magnesium-based alloy.

[0251] Construct a similarity calculation model: Data preparation: Collect and organize magnesium-based alloy preparation data, including various parameters, conditions and results in the historical preparation process.

[0252] Feature selection: Analyze the preparation data of magnesium-based alloys and select features that have an important impact on the similarity calculation, such as temperature, pressure, time, composition ratio, etc.

[0253] Model construction: Based on the selected features, a similarity calculation model is constructed. The model can use similarity measurement methods such as cosine similarity, Euclidean distance, Manhattan distance, etc., or combine multiple methods for comprehensive evaluation.

[0254] Historical data feature vector extraction: From the historical data of magnesium-based alloy preparation, the selected features are extracted for each data to form feature vectors. These feature vectors will be used for subsequent similarity calculations.

[0255] Recommendation parameter feature vector extraction: Similarly, selected features are extracted from the recommendation parameters to form a feature vector of the recommendation parameters.

[0256] Calculate the similarity score: Input feature vector: Input the feature vector of the recommended parameters and the feature vector of the historical data of magnesium-based alloy preparation into the similarity calculation model.

[0257] Similarity calculation: The model calculates the similarity score between the recommended parameters and each piece of historical data based on the input feature vector using a preset similarity measurement method.

[0258] Sorting and selecting matching data: Score sorting: The historical data of magnesium-based alloy preparation are sorted according to the calculated similarity score. The higher the score, the higher the similarity.

[0259] Select matching data: select several pieces of data (such as the first three or five pieces) that match the recommended parameters from the sorted data. The technical solution of the present application selects the data ranked first as the candidate similar prepared data, that is, the data to be processed.

[0260] Retrieve matching preparation data: Data retrieval: Based on the selected matching data, that is, the data to be processed, the corresponding detailed preparation data is retrieved from the historical data of magnesium-based alloy preparation, including specific process parameters, equipment settings, preparation process records, etc.

[0261] Output result: The magnesium-based alloy preparation data that matches the retrieved recommended parameters is used as the data to be processed. The preset database includes the data generated by the magnesium-based alloy preparation process and the corresponding magnesium-based alloy finished product information. The corresponding magnesium-based alloy finished product information is obtained by searching the data to be processed in the database, that is, the magnesium alloy finished product prepared using the parameters and other information in the data to be processed is searched in the database, and the magnesium-based alloy finished product information corresponding to the data to be processed is used as the recommended magnesium-based alloy.

[0262] Specifically, the quality rating and recommendation method of magnesium-based alloys according to the present invention, the step S108, further includes: Receiving and setting the initial conditions of the predator optimization algorithm, the initial conditions of the predator optimization algorithm include initializing the population, determining the operating parameters of the algorithm, substituting each set of data in the initialized population into a preset magnesium-based alloy preparation model, calculating the simulation results corresponding to each set of data in the initialized population, and evaluating the fitness of each set of data according to a preset fitness function to obtain the fitness evaluation result of the initial population; According to the fitness evaluation results of the initial population, the individual with the highest fitness evaluation result of the initial population is selected as the prey, and the individual with the lowest fitness evaluation result of the initial population is selected as the predator. The predator updates its parameters by imitating the parameters of the prey and outputs the updated population data; Substituting the updated population data into the preset magnesium-based alloy preparation model again, calculating new simulation results, and evaluating the fitness of each group of data in the new simulation results to obtain the fitness evaluation results of the updated population; Determine whether the termination condition is met: Check whether the fitness of the fitness evaluation result of the updated population meets the termination condition. If so, output the optimization result; If not satisfied, continue to perform the predation operation; If the judgment result is to continue iteration, the population data will continue to be updated, and the fitness of the new population data will be evaluated. The population data and fitness evaluation results after each round of iteration will be output until the termination condition is met; The individual with the highest fitness is selected from the final population as the optimized simulation data for magnesium-based alloy preparation.

[0263] In step S108, the magnesium-based alloy quality rating and recommendation method of the present invention involves optimizing the magnesium-based alloy preparation simulation data using a predatory optimization algorithm.

[0264] Accept and set the initial conditions for the predator optimization algorithm: Initialize the population: randomly generate a set of initial solutions (i.e., population), each solution represents a set of parameter combinations for the preparation of magnesium-based alloys.

[0265] Determine operating parameters: Set the operating parameters of the predation optimization algorithm, such as the number of iterations, population size, fitness threshold, etc.

[0266] Substitution into model calculation: Substitute each set of data in the initialized population into the preset magnesium-based alloy preparation model, and calculate the simulation results corresponding to each set of data (such as the predicted value of alloy performance).

[0267] Evaluate fitness: According to the preset fitness function, evaluate the fitness of each set of data and obtain the fitness evaluation result of the initial population.

[0268] Choose your prey and predator: Select prey: From the fitness evaluation results of the initial population, select the individual with the highest fitness as the prey, indicating that its parameter combination is better.

[0269] Select predator: At the same time, the individual with the lowest fitness is selected as the predator, indicating that its parameter combination needs to be improved.

[0270] Predator parameters update: Imitating prey parameters: The predator updates its parameters by imitating the parameters of its prey. A certain learning rate or update rule can be used to adjust the predator's parameters.

[0271] Output updated population data: After updating the predator's parameters, new population data is obtained.

[0272] Recalculate the fitness: Substitute the model to calculate new results: Substitute the updated population data into the preset magnesium-based alloy preparation model again to calculate new simulation results.

[0273] Evaluate new fitness: According to the preset fitness function, evaluate the fitness of each set of data in the new simulation results to obtain the fitness evaluation results of the updated population.

[0274] Determine whether the termination condition is met: Check fitness: Check whether the fitness of the current population meets the preset termination conditions (such as reaching the maximum number of iterations, fitness reaching the threshold, etc.).

[0275] Decide on next steps: If the termination condition is met, the optimization result is output, that is, the individual with the highest fitness in the current population is used as the optimized magnesium-based alloy preparation simulation data.

[0276] If the termination condition is not met, the predation operation continues.

[0277] Continue iterating: Update population data: If the judgment result is to continue iteration, the population data will continue to be updated according to the rules of the predation optimization algorithm (such as through selection, crossover, mutation and other operations).

[0278] Evaluate the fitness of the new population: Evaluate the fitness of the updated population data, and output the population data and fitness evaluation results after each round of iteration.

[0279] Repeat iteration: Repeat the above iterative process until the termination condition is met.

[0280] Select the best individual: Select the individual with the highest fitness: From the final population, select the individual with the highest fitness as the optimized magnesium-based alloy preparation simulation data.

[0281] Output optimization results: Output the optimized magnesium-based alloy preparation simulation data for subsequent magnesium-based alloy preparation process or quality rating.

[0282] The technical solution of the present invention effectively solves the problem that the performance test data of the same batch of alloys fluctuates due to fluctuations in raw material composition from different sources and differences in test equipment, test conditions and operator skill levels, thereby affecting the accuracy of quality rating. The following are the specific steps of the technical solution: Obtain the historical tasks of magnesium-based alloy preparation, raw material composition data, performance test data, test condition data, test equipment information and operator skill level data to form a complete magnesium-based alloy data set. Preprocess the data set, including data cleaning, outlier processing, data consistency verification and standardization.

[0283] Key features are extracted from the preprocessed data set, including preparation history task features, raw material composition features, performance test features, test condition features, test equipment features, and operator skill level features. The extracted features are integrated to obtain statistical analysis results of magnesium-based alloy data, providing a basis for subsequent model training.

[0284] Using the statistical analysis results as input, the neural network model is trained to establish a quality rating model for magnesium-based alloys. The model can identify the complex relationship between raw material composition, test conditions and alloy quality, providing an accurate basis for quality rating.

[0285] Substitute the preparation history data into the quality rating model to obtain the quality data set. Collect the composition data of the raw materials of the magnesium-based alloy to be analyzed, and match the corresponding quality level in the quality data set. Retrieve the preparation task corresponding to the quality level, substitute it into the preparation simulation model, and obtain the simulation parameters. Compare the simulation parameters with the real-time quality level to obtain the preparation parameter error data, which provides direction for subsequent optimization.

[0286] The error data is matched with the preset error range knowledge base, and the data exceeding the standard error value is marked as the first type of label, and the data not exceeding the standard error value is marked as the second type of label. The simulation data is optimized using the predatory optimization algorithm to obtain the optimized preparation simulation data. The optimized simulation parameters are matched with the real-time quality level to obtain the matched preparation parameter data as the recommended parameters to improve the accuracy and efficiency of the preparation.

[0287] The recommended parameters are matched in the historical preparation data to find the matching preparation data. The magnesium-based alloy prepared using the matching preparation data is retrieved as the recommended magnesium-based alloy to provide a reliable reference for actual production.

[0288] Through the above steps, the technical solution of the present invention can effectively cope with the challenges brought about by fluctuations in raw material composition, differences in testing equipment, testing conditions and operator skill levels, and significantly improve the accuracy of magnesium-based alloy quality rating.

[0289] Specific embodiments of the present invention: Extract the historical task records of magnesium-based alloy preparation from the company's internal production management system; obtain raw material composition data from the raw material supplier's database; obtain performance test reports from the laboratory; record test environment conditions; obtain test equipment information from the equipment management system; and collect operator information through the human resources management system. Integrate data from different sources to form a comprehensive data set. Perform missing value processing, outlier detection and processing, data consistency verification, data standardization, and data segmentation.

[0290] Extract the characteristics of preparation history tasks, raw material composition, performance test, test condition, test equipment and operator skill level. Integrate the extracted features to form a magnesium-based alloy data set, and perform cleaning, preprocessing and statistical analysis.

[0291] The statistical analysis results are used as the input of the neural network model, standardized, and the data set is divided. The appropriate neural network architecture is selected, the number of network layers and nodes is determined, the activation function, loss function, and optimization algorithm are selected, and the model is trained. The trained model is encapsulated as a quality rating model for magnesium-based alloys, and the relevant data is received as input to output the quality rating results.

[0292] Extract data from the preset database, clean, format, and normalize it, extract the required features, form a feature set, substitute it into the quality rating model to calculate the quality rating results, and form a quality data set. Test the magnesium-based alloy samples to be analyzed, collect raw material composition data, match it with the quality data set, calculate the similarity, and determine the real-time quality level.

[0293] According to the real-time quality level, query and retrieve the preparation task records, substitute them into the preparation simulation model, output the simulation parameters, compare them with the real-time quality level, and obtain the preparation parameter error data. Match the error data with the preset error range knowledge base, classify and mark them, and form a dynamic label data set.

[0294] Substitute the dynamic label data set into the preparation model, apply the predator optimization algorithm to optimize the simulated data, obtain the optimized preparation parameters, match them with the data corresponding to the real-time quality level, and obtain the recommended parameters. Build a similarity calculation model, calculate the similarity score, select the matching historical preparation data as the preparation data matched by the recommended parameters, and retrieve the magnesium-based alloy prepared using the data as the recommended magnesium-based alloy.

[0295] This embodiment solves the impact of raw material composition fluctuations, test condition differences and other issues on the accuracy of magnesium-based alloy quality rating, and achieves effective rating and recommendation of magnesium-based alloy quality.

[0296] The model and construction steps involved in the technical solution of the present invention are as follows: Neural network model: The neural network model is used to establish a quality rating model for magnesium-based alloys and identify the relationship between raw material composition, test conditions and alloy quality.

[0297] Steps for building a neural network model: Extract key features from the preprocessed magnesium-based alloy data set, including raw material composition features, performance test features, test condition features, test equipment features, and operator skill level features. Standardize the extracted features to ensure that the dimensions of different features are consistent. Initialize the model and select a suitable neural network architecture, such as a multi-layer perceptron (MLP). Determine the number of network layers and nodes, such as a three-layer MLP structure, where the number of input layer nodes is equal to the number of features, the number of hidden layer nodes is determined based on experience or experiments, and the number of output layer nodes is the number of quality rating categories. Select an activation function (such as ReLU), a loss function (such as the cross entropy loss function), and an optimization algorithm (such as Adam).

[0298] Randomly initialize the weights and bias parameters of the neural network. For model training, divide the standardized data into training set, validation set, and test set. Use the training set to train the initialized neural network model, calculate the output through forward propagation, calculate the error based on the loss function, and update the model parameters through back propagation.

[0299] The training process is iterated continuously until the preset training rounds are reached or the performance on the validation set no longer improves. During the training process, the early stopping method is used to prevent overfitting, and the model architecture, hyperparameters or regularization are adjusted according to the performance on the validation set. The trained neural network model is encapsulated as a quality rating model for magnesium-based alloys. The model can receive raw material composition data, performance test data, test condition data, test equipment information and operator skill level data of magnesium-based alloys as input, and output the quality rating results of magnesium-based alloys.

[0300] Magnesium-based alloy preparation simulation model: The magnesium-based alloy preparation simulation model is used to predict the performance parameters of magnesium-based alloys (such as tensile strength, yield strength, elongation, hardness, etc.) according to the input preparation task parameters (such as alloy composition, preparation process, test conditions, etc.).

[0301] Steps for constructing a simulation model for magnesium-based alloy preparation: Collect historical data on magnesium-based alloy preparation, including preparation task ID, alloy composition, preparation process parameters, test conditions and corresponding performance test results. Clean, format and normalize the data. Select a suitable model based on data characteristics and problem requirements, such as a physical model, numerical model or machine learning model (such as a neural network, support vector machine, etc.). Design the model structure, determine the input features, output parameters and model parameters. Use historical data on magnesium-based alloy preparation to train the model and adjust the model parameters to improve prediction accuracy. Use a validation set to evaluate model performance and ensure the generalization ability of the model on different data sets. Apply the trained model to the simulation of the actual preparation task and output the predicted performance parameters. Optimize and adjust the model based on feedback from actual applications.

[0302] Similarity calculation model: The similarity calculation model is used to calculate the similarity between the recommended parameters and the historical data of magnesium-based alloy preparation, and find matching preparation data.

[0303] The steps of constructing the similarity calculation model are as follows: Extract key features from the historical data of magnesium-based alloy preparation to form a feature vector. Extract corresponding features from the recommended parameters to form a feature vector of the recommended parameters. Select a suitable similarity measurement method, such as cosine similarity, Euclidean distance, Manhattan distance, etc. Input the feature vector of the recommended parameters and the feature vector of the historical data of magnesium-based alloy preparation into the similarity calculation model. Calculate the similarity score according to the selected similarity measurement method. Sort the historical data of magnesium-based alloy preparation according to the similarity score. Select several pieces of data with the highest similarity as the preparation data matched with the recommended parameters.

[0304] Through the construction and application of the above-mentioned model, the technical solution of the present invention realizes the effective rating and recommendation of the quality of magnesium-based alloys, improves the accuracy of the quality rating of magnesium-based alloys, and provides optimized preparation parameter recommendations.

Claims

1. A quality rating and recommendation method for magnesium-based alloys, characterized in that: include: Step S101, obtaining magnesium-based alloy preparation history tasks, magnesium-based alloy raw material composition data, magnesium-based alloy performance test data, magnesium-based alloy test condition data, magnesium-based alloy test equipment information, and magnesium-based alloy preparation operator skill level data to obtain a magnesium-based alloy data set, and preprocessing the magnesium-based alloy data set to obtain a preprocessed magnesium-based alloy data set; Step S102, extracting features from the preprocessed magnesium-based alloy data set to obtain magnesium-based alloy preparation history task features and magnesium-based alloy data features, the magnesium-based alloy data features include magnesium-based alloy data raw material composition features, magnesium-based alloy data performance test features, magnesium-based alloy data test condition features, magnesium-based alloy data test equipment features, and magnesium-based alloy data operator skill level features, and integrating the magnesium-based alloy data features to obtain magnesium-based alloy data statistical analysis results; Step S103, based on the magnesium-based alloy data characteristics and the magnesium-based alloy data statistical analysis results, a neural network model is used for training to establish a quality rating model for the magnesium-based alloy, where the quality rating model for the magnesium-based alloy is used to identify the relationship between raw material composition, test conditions and alloy quality; Step S104, obtaining historical data of magnesium-based alloy preparation, substituting the historical data of magnesium-based alloy preparation into a quality rating model of magnesium-based alloy, and obtaining a quality data set of magnesium-based alloy; Step S105, collecting raw material composition data in the magnesium-based alloy to be analyzed, matching the raw material composition data in the magnesium-based alloy to be analyzed with the quality data set of the magnesium-based alloy, obtaining the magnesium-based alloy quality level corresponding to the raw material composition data in the magnesium-based alloy to be analyzed, and using the magnesium-based alloy quality level corresponding to the raw material composition data in the magnesium-based alloy to be analyzed as the real-time magnesium-based alloy quality level; Step S106, calling up the magnesium-based alloy preparation task corresponding to the real-time magnesium-based alloy quality level, taking the magnesium-based alloy preparation task corresponding to the real-time magnesium-based alloy quality level as the magnesium-based alloy preparation task to be simulated, substituting the magnesium-based alloy preparation task to be simulated into a preset magnesium-based alloy preparation simulation model, obtaining simulation parameters of the magnesium-based alloy preparation task to be simulated, comparing the simulation parameters of the magnesium-based alloy preparation task to be simulated with the real-time magnesium-based alloy quality level, and obtaining magnesium-based alloy preparation parameter error data; Step S107, matching the magnesium-based alloy preparation error data with a preset magnesium-based alloy preparation error range knowledge base to obtain a magnesium-based alloy preparation error result. If the magnesium-based alloy preparation error result exceeds the standard error value, then retrieving the data exceeding the standard error value in the magnesium-based alloy preparation error data, marking the data exceeding the standard error value as a first type of label, marking the data not exceeding the standard error value as a second type of label, and aggregating the data corresponding to the first type of label and the second type of label to obtain a dynamic label data set; Step S108, substitute the dynamic tag data set into the preset magnesium-based alloy preparation model, output magnesium-based alloy preparation simulation data, optimize the magnesium-based alloy preparation simulation data based on the use of the predator optimization algorithm of the preset magnesium-based alloy preparation model to obtain optimized magnesium-based alloy preparation simulation data, and perform data processing on the optimized magnesium-based alloy preparation simulation data through the preset magnesium-based alloy production model to obtain optimized magnesium-based alloy preparation simulation parameters, match the optimized magnesium-based alloy preparation simulation parameters with the data corresponding to the real-time magnesium-based alloy quality level, obtain magnesium-based alloy preparation parameter data matching the real-time magnesium-based alloy quality level, and use the magnesium-based alloy preparation parameter data matching the real-time magnesium-based alloy quality level as recommended parameters; Step S109, matching the recommended parameters with the historical data of magnesium-based alloy preparation, and using the magnesium-based alloy prepared by using the recommended parameters as the recommended magnesium-based alloy.

2. The quality rating and recommendation method of magnesium-based alloy according to claim 1, characterized in that: The step S102 includes: The characteristics of the magnesium-based alloy preparation history task include the magnesium-based alloy preparation task ID, magnesium-based alloy preparation date, magnesium-based alloy preparation batch, and magnesium-based alloy preparation process parameters; The raw material composition characteristics of magnesium-based alloy data include the content characteristics of each element of magnesium-based alloy data; The data performance test characteristics of magnesium-based alloys include magnesium-based alloy tensile strength characteristics, magnesium-based alloy yield strength characteristics, magnesium-based alloy elongation characteristics and magnesium-based alloy hardness characteristics; The test condition characteristics of magnesium-based alloy data include the test temperature characteristics of magnesium-based alloy data, the test humidity characteristics of magnesium-based alloy data, and the test loading rate characteristics of magnesium-based alloy data; The characteristics of the magnesium-based alloy data testing equipment include the characteristics of the magnesium-based alloy data testing equipment model, the characteristics of the magnesium-based alloy data testing specification, and the characteristics of the magnesium-based alloy data testing calibration record; The skill level characteristics of magnesium-based alloy data operators include the training record characteristics of magnesium-based alloy data operators, the years of experience characteristics of magnesium-based alloy data operators, and the historical operation accuracy characteristics of magnesium-based alloy data operators; The statistical analysis results of magnesium-based alloy data are obtained by integrating the characteristics of historical tasks in magnesium-based alloy preparation, raw material composition characteristics of magnesium-based alloy data, performance test characteristics of magnesium-based alloy data, test condition characteristics of magnesium-based alloy data, test equipment characteristics of magnesium-based alloy data and skill level characteristics of magnesium-based alloy data operators.

3. The quality rating and recommendation method of magnesium-based alloy according to claim 1, characterized in that: The step S103 includes: Using the statistical analysis results of magnesium-based alloy data as the input of the neural network model, initializing the parameters of the neural network model, and obtaining the initialized neural network model; The statistical analysis results of magnesium-based alloy data are divided into a training set and a validation set. The initialized neural network model is trained using the training set. The output is calculated through forward propagation, the error is calculated according to the loss function, and the model parameters are updated through back propagation. The training process is continuously iterated until the preset training rounds are reached or the performance on the validation set is no longer improved, then the trained neural network model is output; The trained neural network model is encapsulated as a quality rating model for magnesium-based alloys. The raw material composition data of magnesium-based alloys, performance test data of magnesium-based alloys, test condition data of magnesium-based alloys, test equipment information of magnesium-based alloys, and skill level data of magnesium-based alloy preparation operators are used as input, and the rating results of the quality of magnesium-based alloys are used as output.

4. The quality rating and recommendation method of magnesium-based alloy according to claim 1, characterized in that: The step S104 includes: Obtaining historical preparation data of magnesium-based alloys from a preset database, the historical preparation data of magnesium-based alloys including historical magnesium-based alloy raw material composition, historical magnesium-based alloy test conditions, historical magnesium-based alloy test equipment information, and historical magnesium-based alloy operator records during the preparation process; Preprocessing the preparation history data of the magnesium-based alloy to obtain the preprocessed preparation history data of the magnesium-based alloy; According to the feature requirements of the magnesium-based alloy quality rating model, the features required by the magnesium-based alloy quality rating model are extracted from the pre-processed magnesium-based alloy preparation history data to obtain an extracted magnesium-based alloy data feature set. The extracted magnesium-based alloy data feature set includes the features required by the magnesium-based alloy quality rating model, including magnesium-based alloy raw material composition data features, magnesium-based alloy performance test data features, magnesium-based alloy test condition data features, magnesium-based alloy test equipment information features, and magnesium-based alloy preparation operator skill level data features; The extracted magnesium-based alloy data feature set is input into the magnesium-based alloy quality rating model. The magnesium-based alloy quality rating model calculates the quality rating of the magnesium-based alloy through the neural network structure inside the magnesium-based alloy quality rating model according to the input data. The quality rating data of the magnesium-based alloy includes the quality score of the magnesium-based alloy, the quality grade of the magnesium-based alloy and the quality classification result of the magnesium-based alloy; The quality rating data of the magnesium-based alloy output by the magnesium-based alloy quality rating model is sorted out, and the quality data of the magnesium-based alloy is associated with the original magnesium-based alloy preparation history data to obtain a quality data set of the magnesium-based alloy.

5. The quality rating and recommendation method of magnesium-based alloy according to claim 1, characterized in that: The step S105 includes: Use magnesium-based alloy raw material composition testing equipment to test the magnesium-based alloy sample to be analyzed, and collect raw material composition data of the magnesium-based alloy to be analyzed; Preprocessing the raw material composition data of the magnesium-based alloy to be analyzed, matching the preprocessed raw material composition data of the magnesium-based alloy to be analyzed with the data in the magnesium-based alloy quality data set, and obtaining historical data records matching the raw material composition of the magnesium-based alloy to be analyzed based on a preset similarity calculation algorithm; Extracting historical data of quality level of magnesium-based alloy from historical data records matching the raw material composition of magnesium-based alloy to be analyzed, and obtaining the quality level of magnesium-based alloy corresponding to the raw material composition data in magnesium-based alloy to be analyzed; The magnesium-based alloy quality level corresponding to the raw material composition data in the magnesium-based alloy to be analyzed is used as the real-time magnesium-based alloy quality level of the magnesium-based alloy to be analyzed.

6. The quality rating and recommendation method of magnesium-based alloy according to claim 1, characterized in that: The step S106 includes: According to the real-time magnesium-based alloy quality level, query and retrieve the magnesium-based alloy preparation task record corresponding to the real-time magnesium-based alloy quality level in a preset database to obtain the magnesium-based alloy preparation task corresponding to the real-time magnesium-based alloy quality level; The magnesium-based alloy preparation task corresponding to the real-time magnesium-based alloy quality level is determined as the magnesium-based alloy preparation task to be simulated, and the magnesium-based alloy preparation task to be simulated is substituted into a preset magnesium-based alloy preparation simulation model, and the preset magnesium-based alloy preparation simulation model outputs simulation parameters of the magnesium-based alloy preparation task to be simulated, and the simulation parameters of the magnesium-based alloy preparation task to be simulated include the expected alloy properties of magnesium-based alloy preparation, variables in the magnesium-based alloy preparation process, control parameters of magnesium-based alloy preparation equipment, and content parameters of each component in magnesium-based alloy preparation; The simulation parameters of the magnesium-based alloy preparation task to be simulated are compared with the real-time magnesium-based alloy quality level to obtain the comparison result of the simulation parameters of the magnesium-based alloy preparation task to be simulated and the real-time magnesium-based alloy quality level, that is, the magnesium-based alloy preparation parameter error data. The magnesium-based alloy preparation parameter error data includes the magnesium-based alloy preparation error value, the magnesium-based alloy preparation error range and the magnesium-based alloy preparation error source analysis.

7. The quality rating and recommendation method of magnesium-based alloy according to claim 1, characterized in that: The step S107 includes: Matching each error index in the magnesium-based alloy preparation error data with the standard error range in the preset error range knowledge base one by one, and obtaining a matching result, i.e., a magnesium-based alloy preparation error result, wherein the magnesium-based alloy preparation error result includes each magnesium-based alloy preparation error index; Determine whether there are indicators exceeding the standard error value for each magnesium-based alloy preparation error index in the magnesium-based alloy preparation error result, and obtain magnesium-based alloy preparation error index data items exceeding the standard error value; Retrieving magnesium-based alloy preparation error index data items exceeding the standard error value from magnesium-based alloy preparation error data; The magnesium-based alloy preparation error index data items exceeding the standard error value are marked as first-class labels, and the magnesium-based alloy preparation error index data items not exceeding the standard error value are marked as second-class labels; The first category label data and the second category label data are aggregated to form a data set including the first category label data and the second category label data, and the data set including the first category label data and the second category label data is used as a dynamic label data set.

8. The quality rating and recommendation method of magnesium-based alloy according to claim 1, characterized in that: The step S108 includes: Substituting the data in the dynamic label data set into a preset magnesium-based alloy preparation model as input parameters of the magnesium-based alloy preparation model, the preset magnesium-based alloy preparation model preliminarily simulates the preparation results according to the input parameters to obtain magnesium-based alloy preparation simulation data; Apply the predatory optimization algorithm to optimize the simulation data of magnesium-based alloy preparation, and output the optimized simulation data of magnesium-based alloy preparation; Substituting the optimized magnesium-based alloy preparation simulation data into a preset magnesium-based alloy production model, and outputting the optimized magnesium-based alloy preparation simulation parameters; Matching the optimized magnesium-based alloy preparation simulation parameters with the data corresponding to the real-time magnesium-based alloy quality level to obtain magnesium-based alloy preparation parameter data matching the real-time magnesium-based alloy quality level; The magnesium-based alloy preparation parameter data that matches the real-time magnesium-based alloy quality level is determined as the recommended parameters, and the recommended parameters are the process parameters for the preparation of the magnesium-based alloy.

9. The quality rating and recommendation method of magnesium-based alloy according to claim 1, characterized in that: The step S109 includes: According to the magnesium-based alloy preparation data, a similarity calculation model is constructed, and the characteristic vector of each data is extracted from the magnesium-based alloy preparation history data to obtain the characteristic vector of magnesium-based alloy preparation, and the characteristic vector of each data is extracted from the recommended parameters to obtain the characteristic vector of the recommended parameters; Using the constructed similarity calculation model, the feature vector of the recommended parameters is compared with the feature vector of the magnesium-based alloy preparation, and the similarity score between the feature vector of the recommended parameters and the feature vector of the magnesium-based alloy preparation is calculated; According to the similarity score between the characteristic vector of the recommended parameter and the characteristic vector of the magnesium-based alloy preparation, the historical data of the magnesium-based alloy preparation are sorted to obtain the sorting result of the historical data of the magnesium-based alloy preparation, and the data ranked first in the sorting result of the historical data of the magnesium-based alloy preparation is used as the data to be processed; The magnesium-based alloy finished product information corresponding to the data to be processed is retrieved from the magnesium-based alloy preparation history data, and the magnesium-based alloy finished product information corresponding to the data to be processed is used as the recommended magnesium-based alloy.

10. The quality rating and recommendation method of magnesium-based alloy according to claim 8, characterized in that: The step S108 further includes: Receiving and setting the initial conditions of the predator optimization algorithm, the initial conditions of the predator optimization algorithm include initializing the population, determining the operating parameters of the algorithm, substituting each set of data in the initialized population into a preset magnesium-based alloy preparation model, calculating the simulation results corresponding to each set of data in the initialized population, and evaluating the fitness of each set of data according to a preset fitness function to obtain the fitness evaluation result of the initial population; According to the fitness evaluation results of the initial population, the individual with the highest fitness evaluation result of the initial population is selected as the prey, and the individual with the lowest fitness evaluation result of the initial population is selected as the predator. The predator updates its parameters by imitating the parameters of the prey and outputs the updated population data; Substituting the updated population data into the preset magnesium-based alloy preparation model again, calculating new simulation results, and evaluating the fitness of each group of data in the new simulation results to obtain the fitness evaluation results of the updated population; Check whether the fitness evaluation result of the updated population meets the termination condition. If so, output the optimization result. If not satisfied, continue to perform the predation operation; If the judgment result is to continue iteration, the population data will continue to be updated, and the fitness of the new population data will be evaluated. The population data and fitness evaluation results after each round of iteration will be output until the termination condition is met; The individual with the highest fitness is selected from the final population as the optimized simulation data for magnesium-based alloy preparation.

Images