Micro-arc oxidation treatment process for producing cosmetic aluminum packaging piece

Abstract

The invention provides a micro-arc oxidation treatment process for production of cosmetic aluminum packaging parts, which comprises the following steps of: judging a synchronous oxidation and coloring state by applying optimized pulse parameters to obtain synchronously completed oxide layer structure data; aiming at the oxide layer structure data, analyzing the complexity of a micropore structure by adopting a neural network model to obtain an input condition of hole sealing treatment; a surface micropore filling rate is obtained according to the input condition, if the filling rate is higher than a preset threshold value, hole sealing operation is executed, and a surface flatness index after hole sealing is determined; judging the temperature and pressure compatibility of texture imprinting according to the surface flatness index to obtain thermal stress distribution data in the imprinting process; and aiming at the thermal stress distribution data, adopting a support vector machine algorithm to optimize imprinting parameters to obtain a texture transfer printing stability index.

Description

Technical Field

[0001] This invention relates to the field of information technology, and in particular to a micro-arc oxidation process for the production of aluminum packaging for cosmetics. Background Technology

[0002] The surface treatment of aluminum packaging components for cosmetics, such as lipstick cases and perfume caps, directly determines the product's appearance, durability, and premium image, holding a crucial position in the highly competitive cosmetics market. Development in this area is not only related to brand differentiation but also requires meeting consumers' stringent demands for a combination of aesthetics and durability.

[0003] Current surface treatment processes mostly employ anodizing combined with dispersed processes such as spraying or electroplating. These processes often rely on chromium passivation or strong acid / alkali pretreatment, making waste emissions unavoidable. Furthermore, multiple independent operations lengthen the production chain, significantly increasing energy consumption and time costs. When pursuing complex color gradients and three-dimensional tactile effects, existing methods struggle to achieve simultaneous oxide layer growth and coloring within the same process, often requiring additional dyeing or coating after oxidation. This not only prolongs the process cycle but also easily affects overall stability due to insufficient interlayer bonding. A deeper technical challenge lies in the difficulty of precisely coordinating the formation of the ceramic layer and the deposition and doping of metal salt ions during micro-arc oxidation. If pulse parameters or electrolyte composition are not properly controlled, the coloring material cannot be uniformly integrated into the plasma channel, resulting in uneven base color distribution or decreased color fastness. This incoordination further restricts the adhesion of subsequent protective layers and textures. Because the microporous structure of the micro-arc oxidation layer surface is complex, poor coloring synchronization weakens the sealing effect, making it difficult for the transparent protective coating to withstand the temperature and pressure of hot stamping after spraying, leading to blurring or peeling of the texture transfer. For example, when mass-producing gradient lipstick cases, if oxidation and coloring cannot be stably completed in the same electrolyte, color differences will appear between products. When hot-pressing textures, the coating may also crack due to unevenness of the underlying layer, resulting in increased scrap rate and delivery delays.

[0004] Therefore, how to achieve simultaneous micro-arc oxidation and coloring while ensuring environmental protection, and how to ensure the continuous, stable, and compatible subsequent sealing, protective coating spraying, and texture imprinting, has become a key issue in improving the efficiency and quality of surface treatment for aluminum cosmetic packaging. Summary of the Invention

[0005] This invention provides a micro-arc oxidation process for the production of aluminum packaging parts for cosmetics, mainly including:

[0006] Pulse parameters and electrolyte composition data during the micro-arc oxidation process are collected. The data is processed using a support vector machine algorithm to obtain a coordination index between the oxide layer growth rate and metal salt ion deposition. The uniform integration degree of the coloring material is obtained based on this coordination index. If the uniform integration degree is lower than a preset threshold, the pulse parameters are adjusted to determine the optimized pulse parameters. The optimized pulse parameters are then applied to determine the state of simultaneous oxidation and coloring, obtaining the oxide layer structure data after simultaneous completion. A neural network model is used to analyze the micropore structure complexity based on the oxide layer structure data to obtain the input conditions for pore sealing. The surface micropore filling rate is obtained based on the input conditions. If the filling rate is higher than a preset threshold, pore sealing is performed, determining the surface smoothness index after sealing. The surface smoothness index is used to determine the temperature and pressure compatibility of texture imprinting, obtaining thermal stress distribution data during the imprinting process. The imprinting parameters are optimized using a support vector machine algorithm based on the thermal stress distribution data to obtain a texture transfer stability index. Furthermore, the step of processing the data using a support vector machine algorithm to obtain a coordination index between the oxide layer growth rate and metal salt ion deposition includes: extracting pulse parameter change features and electrolyte component concentration features from the process dynamic record of the micro-arc oxidation process to determine a classification input set; processing the classification input set using a support vector machine to obtain a matching category between the oxide layer growth rate and metal salt ion deposition; based on the matching category, if the matching category belongs to the coordinated category, the coordination index is judged to be coordinated; if the matching category belongs to the incoherent category, the coordination index is judged to be incoherent. Further, the step of obtaining the uniform integration degree of the coloring substance based on the coordination index, and adjusting the pulse parameters if the uniform integration degree is lower than a preset threshold to determine the optimized pulse parameters includes: collecting temperature fluctuation data through electrolyte temperature monitoring and comparing it with a standard temperature curve to determine the distribution deviation value; for the distribution deviation value, if the uniform integration degree is lower than a preset threshold, using classification feature extraction to separate concentration anomalies from the deviation value to determine the incoherent matching category; adjusting the pulse frequency and voltage amplitude based on the incoherent matching category to obtain an oxide film uniformity index; and determining the optimized pulse parameters based on the oxide film uniformity index. Furthermore, the application of the optimized pulse parameters to determine the state of simultaneous oxidation and coloring, and to obtain the structural data of the synchronously completed oxide layer, includes: obtaining electrolyte environment control data from the application of the optimized pulse parameters; monitoring and judging coloring synchronization evaluation indicators for the oxidation process to obtain state indicator acquisition results; verifying the completion of synchronization using structural data analysis methods based on the state indicator acquisition results, and determining the layer thickness uniformity judgment value; if the layer thickness uniformity judgment value is lower than a preset threshold, adjusting the parameters to obtain temperature deviation correction information, and judging the concentration anomaly identification situation.Based on the concentration anomaly identification, a uniform integration optimization process is used to process the electrolyte environmental control data to obtain the synchronously completed oxide layer structure data. Further, the micropore structure complexity is analyzed using a neural network model based on the oxide layer structure data to obtain the input conditions for pore sealing treatment. This includes: obtaining micropore distribution characteristics from the oxide layer structure data; using Gaussian filtering to remove noise and obtain purified micropore data; extracting pore density distribution indices from the purified micropore data and calculating the distribution uniformity value using principal component analysis to determine micropore structure parameters; inputting the micropore structure parameters into a feedforward neural network model for processing to obtain a complexity assessment result; adjusting the optimization process based on the complexity assessment result and environmental parameter control data; if the structural uniformity is lower than a preset threshold, incorporating temperature correction information to obtain conditional optimization adjustment values; and generating a pore sealing treatment scheme by fusing the conditional optimization adjustment values ​​with the predicted output results to determine the input conditions. Furthermore, the step of obtaining the surface micropore filling rate based on the input conditions, and performing a sealing operation if the filling rate is higher than a preset threshold, to determine the surface smoothness index after sealing, includes: obtaining the surface micropore filling rate and pore density value before the protective coating is applied based on the input conditions; performing a sealing operation to obtain preliminary surface adjustment data if the filling rate is higher than a preset threshold; and calculating the surface smoothness index after sealing by integrating the preliminary surface adjustment data with the pore density value using a weighted average method to determine the smoothness change. Furthermore, the step of determining the temperature and pressure compatibility of texture imprinting through the surface flatness index and obtaining thermal stress distribution data during the imprinting process includes: obtaining texture imprinting temperature and pressure compatibility data through the surface flatness index; determining the initial imprinting conditions if the compatibility is higher than a preset threshold; calculating the deformation distribution pattern by weighted averaging the temperature and pressure data based on the initial imprinting conditions and the material deformation monitoring values; determining the initial level of thermal stress based on the deformation distribution pattern and obtaining stress field adjustment parameters; using the stress field adjustment parameters to integrate the imprinting process variables and obtaining the thermal stress vector through the finite element method to determine the vector distribution characteristics; extracting the thermal stress distribution data during the imprinting process from the vector distribution characteristics; further, the step of optimizing the imprinting parameters using a support vector machine algorithm based on the thermal stress distribution data to obtain a texture transfer stability index includes: obtaining temperature and pressure deformation data based on the thermal stress distribution data; using a support vector machine algorithm to integrate the thermal stress distribution data to obtain an optimized imprinting parameter adjustment scheme; extracting transfer stability elements from the parameter adjustment scheme; and calculating the texture transfer stability index by integrating the transfer stability elements and the deformation index.

[0007] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects:

[0008] This invention discloses a micro-arc oxidation process for cosmetic aluminum packaging, addressing key technical challenges related to oxide layer growth, color uniformity, and surface texture transfer stability through a data-driven and parameter optimization-integrated approach. First, by collecting pulse parameters and electrolyte composition data, a support vector machine (SVM) algorithm is used for classification processing to obtain a coordination index between oxide layer growth rate and metal salt ion deposition. This allows for the assessment of colorant integration and optimization of pulse parameters, ensuring synchronization between oxidation and coloring. Subsequently, a neural network model is used to analyze the complexity of the oxide layer's microporous structure, determining sealing conditions to improve surface smoothness. Finally, SVM optimization of imprinting parameters achieves high stability in texture transfer. This invention, through multi-algorithm collaboration and dynamic parameter adjustment, significantly improves oxide layer quality, color uniformity, and texture transfer effects in the micro-arc oxidation process, providing an efficient solution for complex surface treatments. Attached Figure Description

[0009] Figure 1 This is a flowchart of the micro-arc oxidation process for the production of aluminum cosmetic packaging parts according to the present invention. Detailed Implementation

[0010] To further understand the content of this invention, a detailed description of the invention is provided in conjunction with the accompanying drawings and embodiments. The specific embodiments described herein are for illustrative purposes only and are not intended to limit the invention. It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings.

[0011] like Figure 1 The micro-arc oxidation process for producing aluminum cosmetic packaging components in this embodiment specifically includes:

[0012] S101. By collecting pulse parameters and electrolyte composition data during the micro-arc oxidation process, and using the support vector machine algorithm to classify and process these data, a coordination index between oxide layer growth rate and metal salt ion deposition is obtained.

[0013] By collecting pulse parameters and electrolyte composition data during the micro-arc oxidation process, a dynamic record of the process is obtained. Pulse parameter variation features and electrolyte composition concentration features are extracted from the dynamic record to determine a classification input set. A support vector machine is used to process the classification input set to obtain the matching category between oxide layer growth rate and metal salt ion deposition. Based on the matching category, if the matching category belongs to the coordinated category, the coordination index is determined to be coordinated; if the matching category belongs to the incoherent category, the coordination index is determined to be incoherent.

[0014] In one implementation, pulse parameters and electrolyte composition data are collected during the micro-arc oxidation process using a dedicated sensor.

[0015] Specifically, pulse parameters include voltage amplitude, current density, and pulse frequency, which directly affect the oxide layer formation process.

[0016] For example, when processing aluminum alloy surfaces, voltage sensors are used to monitor pulse voltage changes in real time, while current probes record current density to ensure that the data accurately reflects the process dynamics. Electrolyte composition data is collected using ion concentration analyzers, such as measuring the concentration of silicates or phosphates, which determine the availability of metal salt ions.

[0017] It should be noted that this acquisition method is adaptable to micro-arc oxidation processes on different metal substrates, ensuring the real-time nature and reliability of data acquisition. Furthermore, after preprocessing, the acquired data is classified using a Support Vector Machine (SVM) algorithm. SVM is a supervised learning model that separates data points of different categories by constructing a hyperplane; here, it is applied to distinguish the coordination between oxide layer growth rate and metal salt ion deposition. Specifically, the pulse parameters and electrolyte composition are first converted into feature vectors, for example, using voltage amplitude and ion concentration as input dimensions. Then, the algorithm uses radial basis functions as the kernel function to calculate the similarity between data points to maximize the interval between categories. This classification process helps identify optimal combinations of process parameters and avoids the problem of uneven oxide layer deposition.

[0018] In one possible implementation, for micro-arc oxidation of titanium alloys, the algorithm can classify combinations of high-frequency pulse data and low-concentration electrolyte data, and output classification labels to indicate the degree of coordination.

[0019] For example, oxide growth rate refers to the increase in oxide film thickness per unit time, typically measured and verified using scanning electron microscopy, while metal salt ion deposition involves the migration and binding of ions from the electrolyte to the substrate surface. The reconciliation index is a value calculated using support vector machine classification results.

[0020] For example, the ratio of classification accuracy to growth rate can be used as an indicator; a higher value indicates a more balanced deposition process. The process of obtaining this indicator includes a data training phase and a testing phase. During training, historical datasets are used to optimize model parameters, ensuring the algorithm's applicability in practical micro-arc oxidation.

[0021] Preferably, in another embodiment, for the micro-arc oxidation process of magnesium alloys, the pulse frequency is adjusted to a range of 50Hz to 200Hz, and the concentration of aluminum salt ions in the electrolyte is monitored. These data are classified using a support vector machine algorithm to obtain a coordination index, such as the degree of matching between the growth rate and the ion deposition rate when the growth rate reaches 0.5 micrometers per minute. This approach demonstrates the flexibility of the technical solution within the same field, adapting to different alloy types without changing the core algorithm.

[0022] Understandably, the classification process of the Support Vector Machine (SVM) algorithm includes the selection of support vectors, which are sample points close to the decision boundary used to define the classification boundary. In micro-arc oxidation data, support vectors may correspond to critical pulse parameters, such as anomalous deposition when the voltage exceeds 400V. Through this mechanism, the algorithm achieves effective separation of data and then calculates the coordination index. The technical effect of this process is to improve the prediction accuracy of oxide layer quality; for example, experiments have shown that when the index value is greater than 0.8, the oxide layer hardness increases by 20%.

[0023] Specifically, in one embodiment, pulse parameter data is first collected, such as setting the pulse width to 100 microseconds, and recording the electrolyte pH value as 8 to 10. Then, a support vector machine model is trained using this data, and hyperparameters are optimized using cross-validation. After classification, a coordination index is calculated using a formula, such as the ratio of growth rate to the amount of deposited ions. This detailed process ensures the specific implementation of the technical features in the claims. Further, in another embodiment targeting micro-arc oxidation of stainless steel substrates, data collected includes a current density of 2 A / dm² and the potassium salt concentration in the electrolyte. After algorithmic classification, the coordination index reflects the balance between growth rate and deposition; for example, an index value of 1 indicates an ideal coordination state. This variety of examples supports the versatility of the technical solution without exceeding the scope of micro-arc oxidation.

[0024] S102. Based on the obtained coordination index, obtain the degree of uniform integration of the coloring material in the plasma channel. If the degree of uniform integration is lower than the preset threshold, adjust the pulse frequency and voltage amplitude to determine the optimized pulse parameters.

[0025] The uniform integration degree of the coloring material in the plasma channel is obtained from the coordination index. Temperature fluctuation data is collected by monitoring the electrolyte temperature and compared with a standard temperature curve to determine the distribution deviation value. If the uniform integration degree is lower than a preset threshold, classification feature extraction is used to separate concentration anomalies from the deviation value to determine the incoordination matching category. Based on the incoordination matching category, the pulse frequency and voltage amplitude are adjusted to obtain the oxide film uniformity index. The optimized pulse parameters are determined based on the oxide film uniformity index.

[0026] In one implementation, the uniform integration degree of the coloring material in the plasma channel is further obtained based on the obtained coordination index. The coordination index is a value obtained from support vector machine classification processing, used to reflect the balance between oxide layer growth and ion deposition. The coloring material typically refers to pigment ions or compounds added to the electrolyte, which are integrated into the oxide layer through a micro-arc discharge process to achieve surface coloring. The plasma channel is a local discharge path formed in micro-arc oxidation, generating high-temperature plasma under high voltage to promote material migration. The uniform integration degree can be evaluated by mapping the coordination index to the coloring material distribution data.

[0027] For example, by correlating the index values ​​with the color uniformity observed under an optical microscope, a quantitative score of the degree of integration can be calculated. This acquisition process ensures the accuracy of the process and enables it to adapt to micro-arc oxidation applications with different coloring requirements.

[0028] Specifically, if the uniform integration is below a preset threshold, the pulse frequency and voltage amplitude need to be adjusted. The preset threshold is an empirical value set based on historical process data; for example, a value of 0.7 represents the minimum acceptable level of integration uniformity. The adjustment process first analyzes the current indicators to identify integration unevenness problems caused by excessively high frequency or insufficient voltage. Then, optimization is achieved by gradually increasing the voltage amplitude or decreasing the pulse frequency, for example, adjusting from an initial 400V to 450V while simultaneously reducing the frequency from 150Hz to 100Hz. This adjustment is based on the physical principle that voltage amplitude affects the energy intensity of the plasma channel, promoting uniform diffusion of the coloring material; and pulse frequency controls the discharge interval, preventing material aggregation caused by localized overheating.

[0029] It should be noted that this mechanism improves the stability of the coloring layer in micro-arc oxidation, for example, reducing color spot defects in aluminum alloy surface treatment.

[0030] In one possible implementation, for the micro-arc oxidation process of titanium alloys, after obtaining the degree of uniform integration, if it is below a threshold, optimized pulse parameters are determined. Specific steps include simulating a material distribution model of the plasma channel, using a coordination index as input, and outputting adjustment suggestions.

[0031] For example, when the integration level is 0.6, the algorithm suggests increasing the voltage amplitude by 10% and decreasing the frequency by 20% to achieve uniform integration. This approach demonstrates the targeted nature of parameter optimization, enabling the handling of coloring challenges for different alloys within the same domain without altering the core evaluation logic.

[0032] Preferably, another embodiment targets magnesium alloy substrates, emphasizing the verification of adjusted parameters. Repeated micro-arc oxidation experiments are used to confirm whether the optimized pulse parameters resulted in an integration level exceeding a threshold.

[0033] For example, the integration level was 0.5 under initial parameters, and reached 0.8 after adjustment, indicating a more uniform distribution of the coloring material in the plasma channel. This verification process included real-time monitoring of channel temperature and ion flow to ensure the effectiveness of the adjustments.

[0034] Understandably, the formation of plasma channels depends on the interaction between the electrolyte and the pulse. Adjusting the parameters can balance the pressure gradient within the channel and promote the uniform integration of materials, thereby achieving reliable surface coloring in the field of micro-arc oxidation.

[0035] For example, in the micro-arc oxidation of stainless steel, obtaining the degree of integration involves combining coordination indicators with spectral analysis data, and if it is below a threshold, the parameters are adjusted to optimize channel dynamics.

[0036] Specifically, voltage amplitude adjustment controls channel width, while frequency variation affects material residence time. The goal of this technique is to improve the adhesion of the coloring layer, for example, by observing improved layer thickness uniformity after optimization and reducing the risk of peeling, without introducing additional complexity.

[0037] S103. By applying the optimized pulse parameters in the electrolyte, the synchronous state of oxidation and coloring is determined, and the structural data of the oxide layer that is completed synchronously are obtained.

[0038] Electrolyte environmental control data is obtained from the pulse parameter application. Coloring synchronization evaluation indicators are assessed based on the oxidation process monitoring, resulting in state indicator acquisition results. Using these state indicator acquisition results, structural data analysis methods are employed to verify synchronization completion and determine the layer thickness uniformity judgment value. If the layer thickness uniformity judgment value is lower than a preset threshold, parameter feedback is adjusted to obtain temperature deviation correction information and assess concentration anomaly identification. Based on the concentration anomaly identification, a uniform integration optimization process is used to process the electrolyte environmental control data to obtain synchronized oxide layer structure data.

[0039] In one implementation, optimized pulse parameters, such as adjusted voltage amplitude and pulse frequency, are first introduced into the micro-arc oxidation system by applying them to the electrolyte. The electrolyte typically contains metal salts and coloring agents, such as titanate solutions used in titanium alloy processing. The application process involves immersing the substrate in the electrolyte and activating the pulsed power supply to form and stabilize the plasma channel. This application ensures initial coordination between oxide layer growth and coloring agent incorporation, promoting uniformity of the surface treatment.

[0040] It should be noted that the optimized parameters are based on previous adjustments to the coordination indicators and can balance the ion migration rates within the channels. Furthermore, the simultaneous occurrence of oxidation and coloring is determined by monitoring real-time process parameters.

[0041] Specifically, the synchronization state reflects the degree of matching between oxide layer thickness growth and colorant distribution, for example, by using a spectrometer to detect changes in ion concentration in the channel. If the oxidation rate leads the colorant incorporation, the state is determined to be asynchronous. The judgment process involves collecting channel temperature data and ion flow rate, and then comparing these values ​​with an ideal synchronization model using a preset algorithm. The ideal synchronization model is a framework based on historical data that defines the colorant density threshold corresponding to each micrometer of oxide layer growth. This framework is constructed by simulating the oxidation reaction rate and the mass diffusion equation to ensure the accuracy of the judgment. In practical operation, for example for aluminum alloy substrates, the synchronization state can be quantified as a fraction. If the fraction exceeds 0.8, it indicates that oxidation and coloring are proceeding synchronously, thereby avoiding layer structure defects.

[0042] Preferably, after determining the synchronization status, obtaining the synchronized oxide layer structure data involves the use of scanning electron microscopy. The oxide layer structure data includes layer thickness, porosity, and color distribution maps, which are extracted from the completed sample.

[0043] For example, in the micro-arc oxidation of titanium alloys, data acquisition begins with fixing the sample, followed by analyzing the cross-section of the layers using microscopic imaging to calculate the average thickness, such as 20 micrometers, and mapping the uniformity of the coloring material. This data helps verify process stability without introducing additional variables.

[0044] For example, in the surface treatment of magnesium alloys, after applying optimized pulse parameters, synchronization can be combined with a voltage feedback mechanism. Specifically, the process involves real-time recording of current fluctuations in the electrolyte; if the fluctuations match the oxidation model, synchronization is confirmed. The obtained data further includes elemental composition analysis, such as obtaining crystal structure information through X-ray diffraction. This approach demonstrates the applicability of the technology in lightweight alloys, ensuring the oxide layer possesses corrosion-resistant properties.

[0045] It is understandable that the oxide layer structure data completed simultaneously in the micro-arc oxidation of stainless steel is obtained by integrating multi-source data.

[0046] For example, combining the judgment state with the channel dynamic model outputs structural data such as layer hardness values ​​and color depth maps. The channel dynamic model is a simulation framework that describes the effect of plasma pressure on material distribution and predicts layer structure evolution by inputting pulse parameters. This framework is explained by its thermodynamic basis, balancing oxidation heat and color diffusion to avoid local inhomogeneities. The data acquisition process emphasizes objective quantification, such as measuring layer thickness uniformity deviations of less than 5%, thereby supporting subsequent process optimization.

[0047] In one possible implementation, to accommodate the versatility of different alloys, the determination of synchronization status can be extended to monitoring pH changes.

[0048] Specifically, the electrolyte pH affects ion activity; if it is synchronized with the oxidation rate, the process is considered complete. The obtained structural data includes the micropore distribution, extracted using computed tomography (CT) scans, revealing the penetration depth of the coloring material within the layer. This technique aims to improve layer adhesion, for example, in aluminum alloy applications, reducing the risk of peeling while maintaining process simplicity.

[0049] One embodiment emphasizes the data verification step. After applying the parameters, if the synchronization status is confirmed, the structural data is calibrated through repeated experiments, such as comparing the initial and optimized layer cross-sectional images. This verification ensures the reliability of the data and achieves consistent surface coloring effects in the field of micro-arc oxidation.

[0050] S104. Based on the obtained oxide layer structure data, a neural network model is used to analyze the complexity of the microporous structure and obtain the input conditions for the sealing treatment.

[0051] Micropore distribution characteristics are obtained from the oxide layer structure data. Gaussian filtering is used to remove noise during data preprocessing. Gaussian filtering applies a weighted average to the data points using convolution kernels to obtain purified micropore data. Pore density distribution indices are extracted from the purified micropore data, and principal component analysis (PCA) is used to calculate the uniformity of distribution. PCA obtains the principal axis variance through eigenvalue decomposition of the covariance matrix, determining the micropore structure parameters. These micropore structure parameters are then processed by a feedforward neural network model. The feedforward neural network model receives parameter values ​​at the input layer, applies ReLU activation to the hidden layer to transform the data, and generates an evaluation score at the output layer to obtain a complexity assessment result. Based on the complexity assessment result, the process is adjusted and optimized using environmental parameters obtained from the oxide layer structure data. The uniformity of the structure is judged; if it is below a preset threshold, temperature correction information is incorporated to obtain conditional optimization adjustment values. These conditional optimization adjustment values ​​are then fused with the predicted output results obtained from the complexity assessment to generate a pore sealing treatment scheme, determining the input condition set.

[0052] In one implementation, a method based on a neural network model is proposed to analyze the complexity of micropore structures using oxide layer structure data obtained during micro-arc oxidation, in order to generate input conditions for pore sealing processing. This method is applicable to surface treatment scenarios for materials such as aluminum alloys and titanium alloys, and aims to optimize subsequent process flows through data analysis. The following describes the technical implementation process in detail, from the logical sequence of data processing to model output.

[0053] Specifically, acquiring oxide layer structure data is fundamental to the entire process. This data typically includes information such as layer thickness, porosity, and micropore distribution, and is collected using equipment such as scanning electron microscopes. After data acquisition, preprocessing is required to remove noise, such as smoothing image data, to ensure the accuracy of subsequent analysis. Based on the above steps, the preprocessed data will then be used as input to a neural network model.

[0054] In one possible implementation, the construction of a neural network model is the core of the technical solution. This model employs a multilayer perceptron structure, aiming to extract key features from oxide layer structure data and assess the complexity of the microporous structure.

[0055] It's important to note that microporous structure complexity refers to the combined effect of micropore distribution density, pore size, and pore depth non-uniformity within the oxide layer. During model training, historical oxide layer data is used as the training set. Input features include porosity and micropore distribution density, and the output is a complexity score. The training objective is to enable the model to recognize patterns of different microporous structures and quantify them into a continuous score range, such as 0 to 1, with higher scores indicating more complex structures. After training, by inputting new oxide layer data, the model can quickly output the corresponding complexity score. The key to this process lies in feature extraction and score mapping, ensuring that the analysis results are highly correlated with actual microporous characteristics.

[0056] Preferably, based on the score of the complexity of the microporous structure, the system will further generate input conditions for the sealing process.

[0057] For example, the input conditions for pore sealing include parameters such as pore sealant concentration, processing time, and temperature. A high complexity score indicates uneven micropore distribution and large pore size, requiring an increase in pore sealant concentration or an extension of processing time to ensure effective sealing. Conversely, a low score allows for appropriate reduction of parameter values ​​to conserve resources. The above condition generation logic is based on preset mapping rules to ensure that the output conditions match the complexity analysis results.

[0058] In one embodiment, when applying the above method to the micro-arc oxidation treatment of aluminum alloy materials, the oxide layer structure data is first collected, revealing a high micropore density. After analysis by a neural network model, the complexity score is 0.85, and the system generates sealing conditions with a higher concentration of sealing agent and a longer processing time. This condition, when applied to subsequent processes, effectively improves the sealing performance of the oxide layer. Furthermore, in the processing scenario of titanium alloy materials, the oxide layer data may show smaller but uniformly distributed micropores. The model analysis outputs a complexity score of 0.3, and the system adjusts the sealing conditions accordingly to a lower concentration of sealing agent and a shorter processing time. This flexible parameter adjustment demonstrates the applicability of the technical solution to different material scenarios.

[0059] Understandably, the above method can also be extended to the surface treatment of magnesium alloys. For magnesium alloy oxide layer data, the model analysis process is similar to the aforementioned one, but the weights of feature extraction are adjusted according to material properties, for example, paying more attention to the impact of pore depth inhomogeneity on complexity. The generated sealing treatment conditions will incorporate the chemical properties of the magnesium alloy to ensure the rationality of the process parameters. In another implementation, the neural network model can introduce additional input features, such as oxide layer surface roughness data, to further improve the accuracy of complexity analysis. By integrating multi-dimensional data, the model's output score can more comprehensively reflect the characteristics of the microporous structure, providing more targeted input conditions for sealing treatment. This approach is suitable for scenarios with high surface quality requirements.

[0060] For example, in the micro-arc oxidation process of precision aluminum alloy components, after comprehensively analyzing roughness and micropore distribution data, the model outputs a complexity score and generates corresponding sealing conditions. This multi-feature analysis method can adapt to complex process requirements and ensure the stability of oxide layer performance.

[0061] In one possible implementation, the system can periodically update the training dataset of the neural network model to incorporate the latest oxide layer structure data, taking into account the characteristics of different materials, thereby improving the model's adaptability and analytical accuracy. Through continuous optimization, the technical solution can achieve wider applications in the field of micro-arc oxidation.

[0062] S105. Based on the input conditions, obtain the surface micropore filling rate before the protective coating is sprayed. If the filling rate is higher than the preset threshold, perform the sealing operation and determine the surface flatness index after sealing.

[0063] Based on the input conditions, the surface micropore filling rate and pore density values ​​before the protective coating is applied are obtained. If the filling rate is higher than a preset threshold, a pore sealing operation is performed to obtain preliminary surface adjustment data. For the preliminary surface adjustment data, the pore density values ​​are integrated and a weighted average method is used to calculate the change in smoothness. The weighted average is based on the ratio of the pore density value to the adjustment data to determine the surface smoothness index after pore sealing.

[0064] In one embodiment, for the oxide layer surface after micro-arc oxidation treatment, the surface micropore filling rate before the protective coating is applied is first obtained based on previously generated sealing input conditions. This process is based on acquiring surface image data using an optical microscope or laser scanning equipment.

[0065] Specifically, the micropore filling rate refers to the proportion of micropores in the oxide layer that are covered by the filling material, which is calculated by analyzing the pixel distribution using image processing software.

[0066] For example, the acquired surface image is binarized, with micropore areas marked as black and filled areas marked as white. The proportion of white pixels to total pixels is then calculated as the fill rate. This calculation ensures the objectivity and repeatability of the data, providing a reliable basis for subsequent judgments. Based on the above acquisition steps, if the fill rate is higher than a preset threshold, a sealing operation is performed.

[0067] It should be noted that the preset threshold is usually set according to the material type. For example, for aluminum alloy surfaces, the threshold can be set to 80% to ensure that the surface is dense enough. The judgment logic compares the calculated fill rate with the threshold; if it exceeds the threshold, the operation is triggered. This threshold setting takes into account the porosity characteristics of the oxide layer, avoiding blindly performing pore sealing.

[0068] Preferably, the sealing operation involves uniformly applying a sealing agent into the surface micropores.

[0069] In one possible implementation, for titanium alloy materials, the process involves preheating the surface to a specific temperature and then immersing it in a sealing agent solution for a period of time.

[0070] Specifically, the sealing agent can be a silicate-based material that fills micropores through capillary action, ensuring no noticeable residue on the surface after the process. Details of this step include controlling the soaking time and solution concentration to match the input conditions.

[0071] In one embodiment, for aluminum alloy surface treatment scenarios, if the fill rate exceeds a threshold after obtaining the fill rate, a sealing operation is performed, and then the surface flatness index after sealing is determined. The flatness index refers to the average value of the surface height difference, which is measured by atomic force microscopy.

[0072] For example, multiple regions are scanned, and the average peak-to-valley height difference is calculated as an indicator to quantify surface smoothness. This process ensures the evaluability of the results. Furthermore, for similar scenarios involving magnesium alloys, the micropore filling rate can be adjusted by incorporating pore distribution data.

[0073] For example, if the fill rate calculation shows that it is higher than the threshold, the sealing operation uses a spraying method to apply the sealing agent, and then the flatness index is determined by a laser level.

[0074] Specifically, the index calculation involves measuring the coordinate deviations of multiple points on the surface and obtaining the standard deviation as the flatness value. This method is adapted to the corrosion susceptibility of magnesium alloys.

[0075] Understandably, in another implementation, fill rate acquisition can incorporate multi-point sampling to improve accuracy.

[0076] For example, data is collected from different areas of the oxide layer, the average fill rate is calculated, and if it exceeds a threshold, sealing is performed. The flatness index is then determined using a roughness tester. This extension ensures the comprehensiveness of the process.

[0077] Specifically, for precision titanium alloy components, the flatness index after sealing can be determined automatically using image analysis software. By comparing surface data before and after sealing and calculating the changes in the index, the continuity of the process is demonstrated.

[0078] S106. By determining the surface flatness index, judge the temperature and pressure compatibility of texture imprinting and obtain the thermal stress distribution data during the imprinting process.

[0079] Using the surface smoothness index, texture imprinting temperature and pressure compatibility data are obtained. If the compatibility is higher than a preset threshold, the initial imprinting conditions are determined. For these initial conditions, the temperature and pressure data are fused and weighted to calculate the material deformation monitoring values, resulting in a deformation distribution pattern. Based on this deformation distribution pattern, the initial level of thermal stress is determined, and stress field adjustment parameters are obtained. Using these stress field adjustment parameters, the thermal stress vector is obtained through the finite element method, integrating the imprinting process variables, and the vector distribution characteristics are determined. Thermal stress distribution data during the imprinting process is extracted from these vector distribution characteristics.

[0080] In one implementation, the temperature and pressure compatibility of texture embossing is determined by identifying surface flatness indicators. This process first assesses whether the surface is suitable for embossing operations based on previously obtained flatness indicator data, such as average height difference.

[0081] Specifically, the smoothness index reflects the uniformity of the surface microstructure. If the index value is below a certain threshold, it indicates that the surface is smooth enough to accommodate higher temperature and pressure settings.

[0082] It should be noted that this judgment logic is achieved by comparing the flatness index with an empirical threshold. For example, for aluminum alloy surfaces, the threshold can be set to 5 micrometers to ensure that excessive deformation does not occur during the imprinting process. This step ensures the adaptability of the operating parameters and avoids blind adjustments. Furthermore, after assessing compatibility, thermal stress distribution data during the imprinting process is obtained. Thermal stress distribution data refers to the stress field distribution within the surface material caused by temperature gradients and pressure during imprinting, calculated using finite element simulation software.

[0083] For example, for titanium alloy materials, after the compatibility judgment is passed, the temperature is set to 200 degrees Celsius and the pressure is 10 MPa, and then the flatness index is input into the simulation model.

[0084] Specifically, the simulation process involves meshing the surface, applying boundary conditions such as thermal conductivity and elastic modulus, and then iteratively calculating the stress values ​​at each node. This method takes into account the thermal expansion characteristics of the material, ensuring that the data accurately reflects the actual imprinting scenario. This data can then be used to assess potential cracking risks.

[0085] In one possible implementation, for the surface treatment of magnesium alloys, the temperature and pressure compatibility of texture imprinting can be adjusted by combining multi-point sampling of the flatness index.

[0086] For example, the average values ​​of indicators from multiple regions of the surface are collected, and if compatible, the process proceeds to obtaining the thermal stress distribution.

[0087] Specifically, the calculation of thermal stress distribution involves considering the contact area of ​​the embossing die and the yield strength of the material. The dynamic embossing process is simulated using software, and a stress contour map is output. This approach is adapted to the low melting point characteristics of magnesium alloys, achieving the goal of parameter optimization.

[0088] Understandably, for precision aluminum alloy parts, obtaining thermal stress distribution data during the embossing process can be used to introduce temperature gradient analysis to improve accuracy.

[0089] In one embodiment, after determining compatibility, the simulation software uses a thermal-structural coupling model, inputs the pressure distribution and heat conduction equations, and calculates the stress peak values ​​of each layer on the surface.

[0090] Specifically, this process involves setting an initial temperature field, then gradually applying pressure and monitoring the propagation of stress from the surface to the interior. This approach demonstrates a systematic approach, ensuring that the data supports subsequent optimization. In another implementation, for similar scenarios involving titanium alloys, when assessing compatibility using surface smoothness indicators, historical data can be referenced for threshold calibration.

[0091] For example, if the flatness index indicates low surface roughness, a higher pressure setting is compatible. Subsequently, thermal stress distribution data is obtained through a combination of experimental verification and simulation.

[0092] Specifically, a small-scale imprint test is first conducted to measure the actual stress, and then the model parameters are calibrated using the data. This extension ensures the reliability of the judgment. Furthermore, for magnesium alloy materials, the acquisition of thermal stress distribution data can be refined by incorporating the imprint speed factor.

[0093] In one embodiment, after the compatibility judgment is passed, a velocity variable is added to the simulation to calculate the dynamic thermal stress change.

[0094] Specifically, by adjusting the pressure application rate and observing the uniformity of stress distribution, this process helps to reduce local overload.

[0095] S107. Based on the obtained thermal stress distribution data, the imprinting parameters are optimized using the support vector machine algorithm to obtain the final texture transfer stability index.

[0096] To address the thermal stress distribution, temperature and pressure deformation data are acquired. A support vector machine algorithm is used to integrate this distribution and optimize the imprinting parameters, resulting in a parameter adjustment scheme. From this scheme, transfer stability elements are extracted, and the texture transfer stability index is calculated by fusing these elements with deformation indices. For the texture transfer stability index, imprinting process variable data is acquired, and the finite element method is used to integrate these indices and determine stability adjustment parameters, resulting in an optimized adjustment scheme. From this optimized scheme, material deformation elements are extracted, and the final transfer compatibility index is calculated by fusing these elements with stress field data.

[0097] In one implementation, the imprinting parameters are optimized using a support vector machine algorithm based on the obtained thermal stress distribution data. This process first inputs the thermal stress distribution data, such as stress peak value and distribution uniformity, into the support vector machine model.

[0098] Specifically, Support Vector Machine (SVM) is a supervised learning algorithm that constructs a hyperplane to separate data points, enabling classification or regression tasks. In optimizing imprinting parameters, it is used to regress patterns, predict the stress effects under different combinations of temperature and pressure, and thus find the parameter set that minimizes stress unevenness.

[0099] It should be noted that the advantage of this algorithm lies in handling high-dimensional data. It can map nonlinear problems to a high-dimensional space for linear separation through kernel functions such as radial basis functions, ensuring the robustness of the optimization results. Furthermore, the optimization process includes data preparation and model training phases.

[0100] For example, for aluminum alloys, key features extracted from thermal stress distribution data, such as maximum stress values ​​and gradient changes, are standardized and input into the model. Support vector machines adjust their parameters by minimizing structural risk, i.e., balancing training error and model complexity.

[0101] Specifically, the algorithm iteratively calculates support vectors, i.e., the data points that are closest to the hyperplane, and uses relaxation variables to tolerate small errors. This step ensures that a compatible combination is found within the range of imprinting temperatures from 150 degrees Celsius to 250 degrees Celsius and pressures from 5 MPa to 15 MPa, avoiding deformation caused by excessive stress.

[0102] In one possible implementation, for the optimization of titanium alloy imprinting, the support vector machine algorithm combined with cross-validation method is used to improve accuracy.

[0103] As is understandable, cross-validation involves dividing the dataset into training and test sets and repeatedly training the model to evaluate its generalization ability.

[0104] Specifically, the thermal stress data is first divided into multiple folds, such as a five-fold verification, with each fold serving as a test set in turn. Then, the algorithm optimizes the objective function, solves for the Lagrange multipliers, and obtains the decision function, which is used to predict the optimal parameters.

[0105] For example, for titanium alloys, the optimized temperature is set to 220 degrees Celsius and the pressure to 12 MPa. This process takes into account the material's thermal conductivity, ensuring that parameter adjustments reduce stress concentration. In another embodiment, for magnesium alloys, when using support vector machines for optimization, a mesh search can be introduced to refine the parameters.

[0106] Specifically, the mesh search calculates the stress distribution score for each point using preset temperature and pressure grid points, and then inputs this score into the model for training. A support vector machine (SVM) is then used to construct a regression model, outputting an optimized path that guides parameter adjustments towards lower stress levels. This method adapts to the low-strength characteristics of magnesium alloys and achieves fine-grained parameter control. Furthermore, the optimized imprinting parameters yield the final texture transfer stability index. This index is derived by quantifying the consistency and durability of the transferred surface texture, for example, by calculating texture depth deviation and adhesion values.

[0107] Specifically, the optimal parameters output by the support vector machine are applied to simulated or actual printing, and the transfer result data is collected. Then, a stability score is generated through statistical methods such as analysis of variance.

[0108] It should be noted that this indicator reflects the overall reliability of the printing process; for example, a score above 80 indicates high stability. This method of obtaining the data ensures logical consistency from thermal stress to the transfer result, supporting subsequent process improvements.

[0109] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A micro-arc oxidation process for the production of aluminum packaging parts for cosmetics, characterized in that, include: Pulse parameters and electrolyte composition data during the micro-arc oxidation process are collected. The data is processed using a support vector machine algorithm to obtain a coordination index between the oxide layer growth rate and metal salt ion deposition. The uniform integration degree of the coloring material is obtained based on the coordination index. If the uniform integration degree is lower than a preset threshold, the pulse parameters are adjusted to determine the optimized pulse parameters. The optimized pulse parameters are then applied to determine the state of simultaneous oxidation and coloring to obtain the structure data of the oxide layer that is completed simultaneously. A neural network model is used to analyze the micropore structure complexity based on the oxide layer structure data to obtain the input conditions for the sealing process. The surface micropore filling rate is obtained based on the input conditions; if the filling rate is higher than a preset threshold, a sealing operation is performed, and the surface smoothness index after sealing is determined. The surface smoothness index is used to determine the temperature and pressure compatibility of the texture imprinting, and thermal stress distribution data during the imprinting process is obtained. A support vector machine algorithm is used to optimize the imprinting parameters based on the thermal stress distribution data to obtain a texture transfer stability index.

2. The micro-arc oxidation process for producing cosmetic aluminum packaging components as described in claim 1, characterized in that, The process of using a support vector machine (SVM) algorithm to process the data and obtain a coordination index between oxide layer growth rate and metal salt ion deposition includes: extracting pulse parameter change features and electrolyte component concentration features from the process dynamic record of the micro-arc oxidation process to determine a classification input set; processing the classification input set using an SVM algorithm to obtain a matching category between oxide layer growth rate and metal salt ion deposition; and determining that the coordination index is coordinated if the matching category belongs to the coordinated category, and incompatible if the matching category belongs to the incompatible category.

3. The micro-arc oxidation process for producing aluminum cosmetic packaging components as described in claim 1, characterized in that, The process of obtaining the uniform integration degree of the coloring substance based on the coordination index, and adjusting the pulse parameters if the uniform integration degree is lower than a preset threshold, to determine the optimized pulse parameters, includes: collecting temperature fluctuation data through electrolyte temperature monitoring and comparing it with a standard temperature curve to determine the distribution deviation value; for the distribution deviation value, if the uniform integration degree is lower than the preset threshold, using classification feature extraction to separate concentration anomalies from the deviation value to determine the incoordination matching category; adjusting the pulse frequency and voltage amplitude according to the incoordination matching category to obtain the oxide film uniformity index; and determining the optimized pulse parameters through the oxide film uniformity index.

4. The micro-arc oxidation process for producing cosmetic aluminum packaging components as described in claim 1, characterized in that, The process of applying the optimized pulse parameters to determine the simultaneous oxidation and coloring states and obtaining the synchronously completed oxide layer structure data includes: obtaining electrolyte environment control data from the optimized pulse parameters; monitoring and judging coloring synchronization evaluation indicators for the oxidation process to obtain state indicator acquisition results; verifying the synchronization completion using structural data analysis methods based on the state indicator acquisition results and determining the layer thickness uniformity judgment value; adjusting parameters to obtain temperature deviation correction information and judging concentration anomaly identification based on the layer thickness uniformity judgment value; and processing the electrolyte environment control data using a uniform integration optimization process based on the concentration anomaly identification results to obtain the synchronously completed oxide layer structure data.

5. The micro-arc oxidation process for producing aluminum cosmetic packaging components as described in claim 1, characterized in that, The process of analyzing the micropore structure complexity using a neural network model on the oxide layer structure data to obtain the input conditions for pore sealing treatment includes: obtaining micropore distribution characteristics from the oxide layer structure data; preprocessing the data by using Gaussian filtering to remove noise and obtain purified micropore data; extracting pore density distribution indices from the purified micropore data and calculating the distribution uniformity value using principal component analysis to determine micropore structure parameters; inputting the micropore structure parameters into a feedforward neural network model for processing to obtain a complexity assessment result; adjusting and optimizing the process based on the complexity assessment result and environmental parameter control data; if the structural uniformity is lower than a preset threshold, incorporating temperature correction information to obtain conditional optimization adjustment values; and generating a pore sealing treatment scheme by fusing the conditional optimization adjustment values ​​with the predicted output results, thus determining the input conditions.

6. The micro-arc oxidation process for producing cosmetic aluminum packaging components as described in claim 1, characterized in that, The process of obtaining the surface micropore filling rate based on the input conditions, and performing a sealing operation if the filling rate is higher than a preset threshold, and determining the surface smoothness index after sealing, includes: obtaining the surface micropore filling rate and pore density value before the protective coating is applied using the input conditions; performing a sealing operation to obtain preliminary surface adjustment data if the filling rate is higher than a preset threshold; and calculating the surface smoothness index after sealing by integrating the preliminary surface adjustment data with the pore density value using a weighted average method to determine the smoothness change.

7. The micro-arc oxidation process for producing cosmetic aluminum packaging components as described in claim 1, characterized in that, The step of determining the temperature and pressure compatibility of texture imprinting through the surface flatness index and obtaining thermal stress distribution data during the imprinting process includes: obtaining temperature and pressure compatibility data for texture imprinting through the surface flatness index; determining the initial imprinting conditions if the compatibility is higher than a preset threshold; calculating the deformation distribution pattern by weighted averaging of the temperature and pressure data based on the initial imprinting conditions and the material deformation monitoring values; determining the initial level of thermal stress based on the deformation distribution pattern and obtaining stress field adjustment parameters; integrating the imprinting process variables using the stress field adjustment parameters and obtaining the thermal stress vector through the finite element method to determine the vector distribution characteristics; and extracting the thermal stress distribution data during the imprinting process from the vector distribution characteristics.

8. The micro-arc oxidation process for producing aluminum cosmetic packaging components as described in claim 1, characterized in that, The step of optimizing the imprinting parameters using a support vector machine algorithm based on the thermal stress distribution data to obtain a texture transfer stability index includes: acquiring temperature and pressure deformation data based on the thermal stress distribution data; using a support vector machine algorithm to integrate the thermal stress distribution data to obtain an optimized imprinting parameter adjustment scheme; extracting transfer stability elements from the parameter adjustment scheme; and calculating the texture transfer stability index by fusing the deformation index with the transfer stability elements.

Images