Method, system, device and medium for regulating aluminum alloy anodization process

Abstract

The application relates to the technical field of process control, and provides an aluminum alloy anodic oxidation process regulation method, a system, equipment and a medium.The regulation method comprises the following steps: obtaining the material grade of a workpiece to be processed and target film layer parameters, and generating initial process parameters; starting an anodic oxidation reaction based on the initial process parameters; collecting multidimensional process state data; inputting the multidimensional process state data into a preset long short-term memory neural network model to obtain deviation adjustment parameters; adjusting process parameters in the anodic oxidation reaction process according to the deviation adjustment parameters, and obtaining film layer parameters after a preset time period, judging whether the film layer parameters reach preset target values, and if yes, generating a termination reaction instruction.The process parameters can be dynamically corrected according to real-time changes in the oxidation process, continuous optimization and self-adaptive adjustment of the process parameters are realized, and the requirements of high-end electronics and other fields on oxidation film consistency and reliability can be met.

Description

Technical Field

[0001] This application relates to the field of process control technology, and in particular to methods, systems, equipment and media for controlling aluminum alloy anodizing processes. Background Technology

[0002] Anodizing of aluminum alloys is a surface treatment process that generates an oxide film on the surface of aluminum alloys through electrochemical methods. The thickness, density, and porosity of the anodized film directly affect the corrosion resistance, wear resistance, and insulation properties of the workpiece; therefore, precise control of the anodizing process is crucial. Traditional anodizing process control mainly employs an open-loop control method with fixed parameters. Operators pre-set the electrolyte formula and power parameters based on process manuals or experience. During the oxidation process, the process status is judged solely by manual periodic inspections and offline sampling. When processing different grades of aluminum alloys or when operating conditions fluctuate, the film thickness deviation often exceeds 5μm. Furthermore, the high agglomeration rate during nanoparticle co-deposition makes it difficult to meet the film consistency requirements of the aerospace industry. Summary of the Invention

[0003] This application aims to improve at least one technical problem in the background art.

[0004] This application provides a method for controlling the anodizing process of aluminum alloy, which includes obtaining the material grade and target film parameters of the workpiece to be processed, and generating initial process parameters; Based on the initial process parameters, a start-up reaction command is generated to initiate the anodizing reaction; Collect multi-dimensional process status data during the anodizing reaction to obtain multi-dimensional process status data; Multi-dimensional process status data are input into a preset long short-term memory neural network model to obtain deviation adjustment parameters; The process parameters during the anodizing reaction are adjusted according to the deviation adjustment parameters, and the anodizing reaction continues based on the adjusted process parameters. After a preset time period, the membrane parameters are acquired, and it is determined whether the membrane parameters have reached the preset target value. If so, a termination reaction command is generated.

[0005] According to some technical solutions of this application, the step of obtaining the material grade and target film parameters of the workpiece to be processed, and generating initial process parameters, specifically includes: Obtain the material grade and target film parameters of the workpiece to be processed; Based on the material grade and target membrane parameters, the corresponding electrolyte formula and pulse power parameters are retrieved from the preset material process parameter database; The obtained electrolyte formula and pulse power supply parameters are used as the initial process parameters.

[0006] According to some technical solutions of this application, the process of collecting multi-dimensional process states during the anodizing reaction to obtain multi-dimensional process state data specifically includes: Electrolyte state data are acquired at a first preset frequency; The membrane state data was acquired at a second preset frequency; Power output status data is collected at a third preset frequency; Electrolyte state data, membrane state data, and power output state data are integrated to obtain multi-dimensional process state data.

[0007] According to some technical solutions of this application, the step of inputting multi-dimensional process state data into a preset long short-term memory neural network model to obtain deviation adjustment parameters specifically includes: Multi-dimensional process status data are input into a pre-set long short-term memory neural network model; Temporal features are extracted from multi-dimensional process state data based on a long short-term memory neural network model to obtain temporal features; The timing characteristics are analyzed and processed to obtain the process state replenishment amount and process state adjustment value; The process condition replenishment amount and process condition adjustment value are integrated to obtain the deviation adjustment parameter.

[0008] According to some technical solutions of this application, the step of adjusting the process parameters in the anodizing reaction process according to the deviation adjustment parameters, and continuing the anodizing reaction based on the adjusted process parameters, specifically includes: The deviation adjustment parameters are extracted to obtain the output waveform adjustment value, output power adjustment value, reaction parameter adjustment value, and component concentration replenishment amount; The output waveform of the pulse power supply is adjusted according to the output waveform adjustment value; The output power of the pulse power supply is adjusted according to the output power adjustment value; The concentration of the electrolyte components is adjusted according to the amount of component replenishment. The reaction parameters of the oxidation tank are adjusted according to the reaction parameter adjustment values; The anodic oxidation reaction continues based on the adjusted output waveform and power of the pulse power supply, the adjusted component concentration of the electrolyte, and the adjusted reaction parameters of the oxidation tank.

[0009] According to some technical solutions of this application, the step of acquiring membrane parameters after a preset time period and determining whether the membrane parameters have reached a preset target value, and if so, generating a termination reaction command, specifically includes: Obtain the current film thickness and current film impedance; Compare the current film thickness with the preset target film thickness range; Compare the current film impedance with a preset impedance threshold; If the current membrane thickness meets the preset target membrane thickness range and the membrane impedance detected three times consecutively is greater than or equal to the preset impedance threshold, then the membrane parameters are determined to have reached the preset target value, and a termination reaction command is generated.

[0010] According to some technical solutions of this application, before acquiring multi-dimensional process states during the anodizing reaction to obtain multi-dimensional process state data, the method further includes: Obtain historical process datasets; Construct an initial long short-term memory neural network model; The initial long short-term memory neural network model is trained using the historical process dataset to obtain a preset long short-term memory neural network model.

[0011] This application also provides a control system for an aluminum alloy anodizing process, comprising: The generation module is used to obtain the material grade and target film parameters of the workpiece to be processed, and to generate initial process parameters; The startup module is used to generate a startup reaction command based on the initial process parameters to initiate the anodizing reaction; The acquisition module is used to collect multi-dimensional process states during the oxidation reaction process to obtain multi-dimensional process state data; The parameter generation module is used to input multi-dimensional process state data into a preset long short-term memory neural network model to obtain deviation adjustment parameters; The adjustment module is used to adjust the process parameters in the oxidation reaction process according to the deviation adjustment parameters, and to continue the anodizing reaction based on the adjusted process parameters; The judgment module is used to acquire membrane parameters after a preset time period and determine whether the membrane parameters have reached the preset target value. If so, a termination reaction command is generated.

[0012] This application also provides a control device for an aluminum alloy anodizing process, the control device comprising: a memory and at least one processor, the memory storing instructions; at least one processor calling the instructions in the memory to cause the control device for the aluminum alloy anodizing process to execute each step of the control method for the aluminum alloy anodizing process as described in any of the above technical solutions.

[0013] This application also provides a computer-readable storage medium storing instructions that, when executed by a processor, implement the various steps of the control method for the aluminum alloy anodizing process described in the above technical solution.

[0014] The method for controlling the anodizing process of aluminum alloys provided in this application has at least the following beneficial effects: by setting up an LSTM model and monitoring multi-dimensional data, the process parameters can be dynamically corrected according to the real-time changes in the oxidation process, realizing continuous optimization and adaptive adjustment of the process parameters. This is beneficial for controlling the film thickness deviation, improving the uniformity of nanoparticle distribution, and eliminating scorching defects in thick films. This is beneficial for meeting the requirements of high-end electronics and other fields for the consistency and reliability of oxide films. At the same time, it can be adapted to various types of aluminum alloys, eliminating the need to frequently change the electrolyte formula according to different grades. Attached Figure Description

[0015] Figure 1 A schematic flowchart illustrating the method for controlling the aluminum alloy anodizing process provided in this application embodiment; Figure 2 This is a structural diagram of the control system for the aluminum alloy anodizing process provided in the embodiments of this application; Figure 3 This is a schematic diagram of the control equipment for the aluminum alloy anodizing process provided in an embodiment of this application. Detailed Implementation

[0016] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0017] In the description of this application, it should be understood that the orientation descriptions, such as up, down, front, back, left, right, etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation or be constructed or operated in a specific orientation. Therefore, they should not be construed as limiting the present invention.

[0018] In the description of this application, unless otherwise expressly defined, terms such as "setup," "installation," and "connection" should be interpreted broadly, and those skilled in the art can reasonably determine the specific meaning of the above terms in this application in conjunction with the specific content of the technical solution.

[0019] The following is combined with Figures 1 to 3 Embodiments of the present invention will be described.

[0020] This application provides a method for controlling the anodizing process of aluminum alloys, comprising: S100: Obtain the material grade and target film parameters of the workpiece to be processed, and generate initial process parameters. For example, the workpiece to be processed is a 7075 aluminum alloy structural part for aerospace. Input the material grade "7075" and the target film parameters as film thickness 30μm, salt spray resistance time 1200h, and porosity 15%. Retrieve the initial process parameters corresponding to 7075 aluminum alloy and target parameters from the preset material process parameter database. These parameters include an electrolyte formula of sulfuric acid 180g / L, oxalic acid 15g / L, and boric acid buffer system 3g / L, and pulse power supply parameters of frequency 200Hz, on / off ratio 1:1.5, peak voltage 85V, and current density 1.2A / dm².

[0021] S200 generates a start-up reaction command based on initial process parameters to initiate the anodizing reaction; The S300 collects multi-dimensional process status data during the anodizing reaction. During the oxidation process, various sensors and instruments deployed in the oxidation tank and equipment continuously collect data at preset frequencies. For example, the temperature sensor collects the electrolyte temperature every second; the pH sensor collects the electrolyte pH value every 5 seconds; the ultrasonic concentration meter collects the sulfuric acid concentration every minute; the Al³⁺ ion-selective electrode collects the aluminum ion concentration every 5 minutes; the in-situ ellipsometry collects the film thickness every 30 seconds; the electrochemical impedance spectroscopy collects the film impedance every 2 minutes; and the Hall sensor collects the pulse frequency, on / off ratio, peak voltage, and current density every 100 ms, forming a multi-dimensional process status dataset.

[0022] S400 inputs multi-dimensional process status data into a preset long short-term memory neural network model to obtain deviation adjustment parameters; for example, after standardizing 12 types of data collected in real time, including temperature, pH value, sulfuric acid concentration, Al³⁺ concentration, pulse frequency, on / off ratio, peak voltage, current density, film thickness, film impedance, film growth rate, and real-time energy consumption, the data is input into a pre-trained LSTM neural network model. The model extracts temporal features through three LSTM hidden layers and performs predictive analysis through fully connected layers, outputting eight-dimensional bias adjustment parameters. These parameters include sulfuric acid supply (0.5L), disodium EDTA supply, pulse frequency adjustment (z), on / off ratio adjustment, peak voltage adjustment, current density adjustment, stirring speed adjustment, and temperature. For example, the output parameters might be: sulfuric acid supply (0.5L), disodium EDTA supply (0.2L), pulse frequency adjustment (decreased by 50Hz), on / off ratio adjustment (increased by 0.3), peak voltage adjustment (decreased by 2V), current density adjustment (decreased by 0.1A / dm²), stirring speed adjustment (increased by 50r / min), and temperature adjustment (decreased by 1℃).

[0023] The S500 adjusts the process parameters during the anodizing reaction based on the deviation adjustment parameters, and continues the anodizing reaction based on the adjusted process parameters. The obtained deviation adjustment parameters are parsed into specific execution commands, such as sending commands to the pulse power supply module to adjust the frequency from 200Hz to 150Hz, the on / off ratio from 1:1.5 to 1:1.8, the peak voltage from 85V to 83V, and the current density from 1.2A / dm² to 1.1A / dm²; sending commands to the metering pump of the intelligent electrolyte system to add 0.5L of sulfuric acid and 0.2L of disodium EDTA; sending commands to the stirrer to increase the rotation speed by 50r / min; and sending commands to the temperature control module to decrease the set temperature by 1℃. After the adjustments are completed, the anodizing reaction continues to operate.

[0024] S600: After a preset time period, the system acquires the membrane parameters and determines whether the membrane parameters have reached the preset target value. If so, a termination command is generated. For example, the preset time period is 30 seconds. Every 30 seconds, the system repeats steps S300 to S500. For example, after 45 minutes of oxidation reaction, the system detects that the current membrane thickness is 30.2 μm, and the membrane impedance reaches the preset threshold of 1.2 × 10⁻⁶ in three consecutive detections. 6 Ω·cm². The system determines that the membrane parameters have reached the preset target value, generates a termination reaction command, gradually reduces the output voltage of the pulse power supply to zero, drains the electrolyte from the oxidation tank, and transfers the workpiece to the post-processing stage.

[0025] In some embodiments, step S100 involves obtaining the material grade and target film parameters of the workpiece to be processed, and generating initial process parameters, specifically including: S110, obtain the material grade and target film parameters of the workpiece to be processed; S120, based on the material grade and target film parameters, retrieves the corresponding electrolyte formula and pulse power parameters from the preset material process parameter database; upon receiving valid input information, it triggers the preset material process parameter database retrieval command. This database stores the process parameter mapping relationship of mainstream aluminum alloy grades such as 1 series, 2 series, 6 series, and 7 series, including electrolyte formulas and pulse power parameter ranges corresponding to different target film parameters and process priorities; S130: The retrieved electrolyte formula and pulse power supply parameters are used as initial process parameters. The electrolyte formula and pulse power supply parameters retrieved from the database are integrated to generate the initial process parameters corresponding to the 2024 aluminum alloy workpiece. Simultaneously, reaction parameters such as the initial temperature and pH value of the oxidation tank are supplemented based on the standard operating conditions of the oxidation tank. This achieves precise matching of process parameters for different aluminum alloy grades and different target film parameters, adapting to the anodizing requirements of multiple aluminum alloy grades. For example, the workpiece to be processed is of grade "2024" and the target film parameters are "15μm / 500h / 10%". The database query results show that the electrolyte formula uses 150g / L sulfuric acid, 8g / L citric acid, and 2g / L boric acid buffer system; the pulse power supply parameters are a frequency of 300Hz, an on / off ratio of 1:1.2, a peak voltage of 75V, and a current density of 1.0A / dm². The retrieved electrolyte formula and pulse power supply parameters are used as the initial process parameters.

[0026] In some embodiments, step S300 involves collecting multi-dimensional process states during the anodizing reaction to obtain multi-dimensional process state data, specifically including: S310, acquire electrolyte state data at a first preset frequency; for example, electrolyte temperature is acquired once every 1 second using a PT100 platinum resistance sensor; pH value is acquired once every 5 seconds using a glass electrode pH meter; sulfuric acid concentration is acquired once every 60 seconds using an ultrasonic concentration meter, with a measurement range of 100-300 g / L; Al³⁺ concentration is acquired once every 300 seconds using an ion-selective electrode, with a measurement range of 0-50 g / L.

[0027] S320: Film state data were acquired at a second preset frequency; film thickness was acquired using an in-situ ellipsometry (JA Woollam M-2000D), measured every 30 seconds, with a measurement range of 190-1700 nm. Film impedance was acquired using an electrochemical impedance spectroscopy (Parstat 4000), measured every 2 minutes, with a test frequency range of 10 mHz-100 kHz.

[0028] S330 collects power output status data at a third preset frequency; pulse frequency, on / off ratio, peak voltage and current density are collected by Hall sensors integrated inside the pulse power supply, once every 100ms.

[0029] S340 integrates electrolyte state data, membrane state data, and power output state data to obtain multi-dimensional process state data. All collected data is aligned and packaged according to a unified timestamp to generate structured data frames. Each data frame contains information such as acquisition time, data category, data value, and data unit, facilitating subsequent model processing.

[0030] In some embodiments, step S400 involves inputting multi-dimensional process state data into a preset long short-term memory neural network model to obtain deviation adjustment parameters, specifically including: S410 inputs multi-dimensional process status data into a pre-set Long Short-Term Memory (LSTM) neural network model; at a certain moment, all multi-dimensional process status data collected within the previous 60 seconds are organized into a 60×12 matrix, which is used as the input to the LSTM model. The 12 feature dimensions are, in order: temperature, pH value, sulfuric acid concentration, Al³⁺ concentration, pulse frequency, on / off ratio, peak voltage, current density, film thickness, film impedance, film growth rate, and real-time energy consumption.

[0031] S420 uses a Long Short-Term Memory (LSTM) neural network model to extract time-series features from multi-dimensional process state data. The LSTM model architecture used in this embodiment consists of a 12-dimensional input layer, a first LSTM layer with 64 neurons, a second LSTM layer with 128 neurons, a third LSTM layer with 64 neurons, followed by a fully connected layer with 128 neurons and a ReLU activation function. The model performs forward propagation on the input data across 60 time steps. Each LSTM layer extracts long-term dependencies from the data through forget gates, input gates, and output gates. After processing by three LSTM layers, the model compresses the time-series data into a 128-dimensional time-series feature vector.

[0032] S430 analyzes and processes the timing features to obtain the process state replenishment amount and process state adjustment value; the timing feature vector is input to the fully connected layer and the output layer. The output layer uses a linear activation function to generate 8 output values. The model calculates the process state replenishment amount and process state adjustment value based on the current process state. For example, the process state replenishment amount is 0.3L of sulfuric acid and 0.1L of disodium EDTA; the process state adjustment value is: pulse frequency adjustment value increased by 30Hz, on / off ratio adjustment value decreased by 0.2, peak voltage adjustment value increased by 3V, current density adjustment value increased by 0.15A / dm², stirring speed adjustment value decreased by 20r / min, and temperature adjustment value increased by 0.5℃.

[0033] S440 integrates the process condition replenishment amount and process condition adjustment value to obtain the deviation adjustment parameter. The eight output values ​​are encapsulated into a deviation adjustment parameter data package according to a preset format, where the data package includes the parameter name, adjustment direction, and adjustment amount.

[0034] In some embodiments, step S500 involves adjusting the process parameters during the anodizing reaction according to the deviation adjustment parameters, and continuing the anodizing reaction based on the adjusted process parameters. Specifically, this includes: S510 extracts the deviation adjustment parameters to obtain the output waveform adjustment value, output power adjustment value, reaction parameter adjustment value, and component concentration replenishment amount; after receiving the deviation adjustment parameter data packet, it parses and classifies it to obtain the output waveform adjustment value, output power adjustment value, reaction parameter adjustment value, and component concentration replenishment amount.

[0035] S520 adjusts the output waveform of the pulse power supply according to the output waveform adjustment value; for example, adjusting the pulse frequency from the current value of 200Hz to 230Hz, and the on / off ratio from the current value of 1:1.5 to 1:1.3. After receiving the command, the pulse power supply executes the new waveform parameters at the beginning of the next pulse cycle.

[0036] In this embodiment, when the LSTM model detects that the film growth rate is lower than the target value, it indicates insufficient kinetics in the anodic oxidation reaction. At this time, the output waveform adjustment value is a negative frequency adjustment value and a positive on / off ratio adjustment value, i.e., reducing the pulse frequency and increasing the on / off ratio. For example, when the film growth rate is lower than the target rate, the pulse frequency is reduced by 30-50 Hz from the current value of 200 Hz to 150-170 Hz, and the on / off ratio is increased from the current value of 1:1.5 to 1:1.8-1:2.0. Reducing the frequency prolongs the duration of each pulse, and increasing the on / off ratio increases the proportion of energized time, thereby enhancing the kinetics of the oxidation reaction and promoting film growth. Conversely, when the film growth rate is higher than the target rate, it indicates that the reaction is too vigorous and may cause the film to burn. At this time, the output waveform needs to increase the pulse frequency and decrease the on / off ratio. For example, the pulse frequency is increased by 30-50 Hz from the current value of 200 Hz to 230-250 Hz, and the on / off ratio is decreased from the current value of 1:1.5 to 1:1.2-1:1.0. Increasing the frequency can increase the number of pulses but shorten the time of a single pulse. Reducing the on / off ratio can reduce the proportion of energized time, thereby suppressing excessively rapid film growth and preventing film defects.

[0037] The S530 adjusts the output power of the pulse power supply according to the output power adjustment value; for example, it sends peak voltage and current density adjustment commands to the pulse power supply, adjusting the peak voltage from 83V to 86V and the current density from 1.1A / dm² to 1.25A / dm². The IGBT module inside the pulse power supply adjusts the conduction angle according to the commands to achieve precise control of the output power.

[0038] In this embodiment, when the LSTM model detects that the film impedance is lower than a preset threshold, it indicates insufficient film density. At this time, the output power adjustment value is a positive peak voltage adjustment value and a positive current density adjustment value, i.e., increasing the peak voltage and current density. For example, when the film impedance is lower than the preset threshold, the peak voltage increases from the current value of 83V by 5-10V to 88-93V, and the current density increases from the current value of 1.1A / dm² by 0.1-0.2A / dm² to 1.2-1.3A / dm². Increasing the voltage enhances the electric field strength, promoting oxide film densification; increasing the current density increases the reaction rate, accelerates film growth, and thus significantly improves film density. Conversely, when the detected real-time energy consumption is higher than a preset energy consumption threshold, the peak voltage and current density can be reduced. Specifically, the peak voltage will be reduced by 3-5V from the current value of 83V to 78-80V, and the current density will be reduced by 0.1-0.15A / dm² from the current value of 1.1A / dm² to 0.95-1.0A / dm². Reducing the voltage and current density can reduce instantaneous power, achieving energy-saving operation while ensuring the quality of the membrane layer.

[0039] S540 adjusts the electrolyte component concentration according to the component concentration replenishment amount; for example, it sends a replenishment command to the metering pump of the electrolyte system. The sulfuric acid metering pump starts and runs at a flow rate of 0.5 L / min for 36 seconds to replenish 0.3 L of sulfuric acid; the EDTA disodium metering pump starts and runs at a flow rate of 0.2 L / min for 30 seconds to replenish 0.1 L of EDTA disodium solution.

[0040] In some optional embodiments, when the Al³⁺ ion-selective electrode detects an aluminum ion concentration exceeding 25 g / L, it indicates that there are too many dissolved aluminum ions in the electrolyte, which will affect the film quality and reaction stability. At this point, the component concentration replenishment includes EDTA disodium replenishment. The system sends a command to the metering pump to inject 0.5-1.0 g / L of EDTA disodium solution at a flow rate of 0.2-0.5 L / min. EDTA disodium complexes with free aluminum ions, reducing the effective aluminum ion concentration and maintaining electrolyte activity.

[0041] In another alternative embodiment, when the pH sensor detects a pH value below 1.2, it indicates that the electrolyte is too acidic, which may lead to excessively rapid membrane dissolution. In this case, the component concentration replenishment includes borax replenishment; the system injects a 0.3-0.8 g / L borax solution at a flow rate of 0.1-0.3 L / min. The borax neutralizes excess acid radicals, causing the pH to rise back to the target range of 1.2-1.8. When the pH sensor detects a pH value above 1.8, it indicates that the electrolyte is insufficiently acidic, which may lead to insufficient membrane growth kinetics. In this case, the component concentration replenishment includes sulfuric acid replenishment; the system injects a 0.5-1.5 g / L sulfuric acid solution at a flow rate of 0.3-0.8 L / min, causing the pH to fall back to the target range.

[0042] S550 adjusts the reaction parameters of the oxidation tank according to the reaction parameter adjustment value; for example, it sends a command to the stirrer drive module to adjust the stirring motor speed from 300 r / min to 280 r / min; and sends a command to the temperature control module to adjust the set temperature of the heating and cooling system from 22.5℃ to 23.0℃.

[0043] In some optional embodiments, when the temperature sensor detects that the electrolyte temperature is higher than the target range, it indicates that excessive exothermic reaction may lead to a loose film. In this case, the reaction parameter adjustment value is a negative temperature adjustment value, and the system sends a cooling command to the temperature control module to lower the set temperature by 1-3°C, initiating a cooling cycle to bring the temperature back to the target range. When the temperature is lower than the lower limit of the target range, it indicates insufficient reaction kinetics. In this case, the reaction parameter adjustment value is a positive temperature adjustment value, and the system sends a heating command to raise the set temperature by 1-3°C, activating the heater to bring the temperature back to the target range. Furthermore, when a decrease in film thickness uniformity or an increase in local defects is detected, it indicates that the electrolyte stirring may be uneven. In this case, the reaction parameter adjustment value adjusts the stirring speed according to the direction of deviation. For example, if the local thickness is too thin, it indicates insufficient ion concentration in that area; therefore, the stirring speed is increased by 20-50 r / min to enhance convection and promote uniform ion distribution. If local porosity defects appear, it indicates that excessive stirring introduces air bubbles; therefore, the stirring speed is decreased by 20-50 r / min to reduce air bubble entrainment.

[0044] S560 continues the anodic oxidation reaction based on the adjusted output waveform and power of the pulse power supply, the adjusted component concentration of the electrolyte, and the adjusted reaction parameters of the oxidation tank. After all adjustment commands are executed, the system enters a stable operating state, and the workpiece continues the oxidation reaction under the adjusted process conditions.

[0045] In some embodiments, before S300, which involves collecting multi-dimensional process states during the oxidation reaction to obtain multi-dimensional process state data, the method further includes: S301, Obtain historical process dataset; Extract anodizing production data accumulated over the past two years from the enterprise database, covering multiple sets of process records for 1-series to 7-series aluminum alloys. Each set of records includes input data and output labels.

[0046] The input data includes electrolyte temperature, pH value, sulfuric acid concentration, Al³⁺ concentration, pulse frequency, on / off ratio, peak voltage, current density, membrane thickness, membrane impedance, membrane growth rate, and real-time energy consumption. The output labels are the optimal adjustment parameters determined based on the final product quality assessment, including sulfuric acid supply, disodium EDTA supply, pulse frequency adjustment, on / off ratio adjustment, peak voltage adjustment, current density adjustment, stirring speed adjustment, and temperature adjustment.

[0047] S302, construct the initial Long Short-Term Memory (LSTM) neural network model; use the TensorFlow deep learning framework to build the LSTM model. The model structure is as follows: input layer dimension 12, first LSTM layer with 64 neurons, second LSTM layer with 128 neurons, third LSTM layer with 64 neurons, fully connected layer with 128 neurons (activated by ReLU), and output layer with 8 neurons (linear activation). The model is compiled using the Adam optimizer with a learning rate of 0.001 and a loss function of mean squared error.

[0048] S303, the initial Long Short-Term Memory (LSTM) neural network model is trained using the historical process dataset to obtain a preset LSM neural network model. Multiple sets of data are divided into training, validation, and test sets in a 7:2:1 ratio. During training, a batch size of 32 and 1000 iterations are used. After each iteration, the model performance is evaluated on the validation set. Training is stopped early when the validation set loss no longer decreases after 10 consecutive iterations. After training, the model's prediction accuracy is evaluated on the test set to be 98.5%, and the MSE loss value is 0.008, meeting the deployment requirements. The trained model parameters are then fixed and deployed to the central intelligent controller for use as a preset model. During production, after every 100 batches of processing, the system automatically extracts 1000 newly generated sets of data for incremental training of the model, achieving online model updates.

[0049] In some embodiments, step S600 involves acquiring membrane parameters after a preset time period and determining whether the membrane parameters have reached a preset target value. If so, a termination reaction command is generated, specifically including: S610, obtain the current film thickness and current film impedance; for example, the current film thickness is acquired as 29.8 μm using an in-situ ellipsometry and the current film impedance is acquired as 1.25 × 10⁻⁶ μm using an electrochemical impedance spectroscopy. 6 Ω·cm².

[0050] S620, compare the current film thickness with the preset target film thickness range; in this embodiment, the preset target film thickness is 30μm, and the allowable deviation range is ±2μm, that is, the target thickness range is 28μm to 32μm. The current film thickness of 29.8μm falls within this range, and the thickness condition meets the requirements.

[0051] S630, compare the current film impedance with a preset impedance threshold; in this embodiment, the preset impedance threshold is 1.2 × 10⁻⁶. 6 Ω·cm², this threshold corresponds to the membrane density meeting the requirement of salt spray resistance for 1200h. The current membrane impedance is 1.25 × 10⁻⁶. 6 The impedance Ω·cm² is greater than the preset threshold, and the impedance condition meets the requirements.

[0052] S640, if the current film thickness meets the preset target film thickness range and the film impedance measured three consecutive times is greater than or equal to the preset impedance threshold, then the film parameters are determined to have reached the preset target value, and a termination command is generated. Based on the previous two detection records, at the 37th minute, the film thickness was 29.5 μm and the impedance was 1.23 × 10⁻⁶. 6 Ω·cm²; at 36.5 minutes, the film thickness was 29.2 μm and the impedance was 1.21 × 10⁻⁶. 6 Ω·cm². The thickness values ​​from all three tests were within the target range, and the impedance values ​​were all greater than the threshold, meeting the requirement of three consecutive successful tests. The system determined that the film parameters had reached the preset target values, generated a termination command, and terminated the anodizing reaction.

[0053] This application also provides a control system for anodizing an aluminum alloy process, which includes a generation module 100, a start-up module 200, a data acquisition module 300, a parameter generation module 400, an adjustment module 500, and a judgment module 600.

[0054] The process includes: a generation module 100 for acquiring the material grade and target film parameters of the workpiece to be processed, and generating initial process parameters; a start module 200 for generating a start reaction command based on the initial process parameters to initiate the anodizing reaction; an acquisition module 300 for acquiring multi-dimensional process states during the anodizing reaction to obtain multi-dimensional process state data; a parameter generation module 400 for inputting the multi-dimensional process state data into a preset long short-term memory neural network model to obtain deviation adjustment parameters; an adjustment module 500 for adjusting the process parameters during the anodizing reaction according to the deviation adjustment parameters, and continuing the anodizing reaction based on the adjusted process parameters; and a judgment module 600 for acquiring film parameters after a preset time period and judging whether the film parameters have reached the preset target value. If so, a termination reaction command is generated.

[0055] Figure 3This is a schematic diagram of the structure of a control device for an aluminum alloy anodizing process provided in an embodiment of the present invention. The control device 700 for the aluminum alloy anodizing process can vary considerably due to different configurations or performance characteristics. It may include one or more central processing units (CPUs) 710 (e.g., one or more processors) and a memory 720, and one or more storage media 730 (e.g., one or more mass storage devices) storing application programs 733 or data 732. The memory 720 and storage media 730 can be temporary or persistent storage. The program stored in the storage media 730 may include one or more modules (not shown in the diagram), each module including a series of instruction operations on the control device 700 for the aluminum alloy anodizing process. Furthermore, the processor 710 may be configured to communicate with the storage media 730 and execute the series of instruction operations in the storage media 730 on the control device 700 for the aluminum alloy anodizing process to implement the steps of the control method for the aluminum alloy anodizing process provided in the above-described method embodiments.

[0056] The control equipment 700 for the aluminum alloy anodizing process may also include one or more power supplies 740, one or more wired or wireless network interfaces 750, one or more input / output interfaces 760, and / or one or more operating systems 731, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. Those skilled in the art will understand that... Figure 3 The control equipment structure shown for the aluminum alloy anodizing process does not constitute a limitation on the control equipment for the aluminum alloy anodizing process. It may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0057] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when executed on a computer, cause the computer to perform the steps of a control method for anodizing an aluminum alloy process.

[0058] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the system, device, or unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0059] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0060] The preferred embodiments of the present invention have been described in detail above, but the present disclosure is not limited to the embodiments described. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention, and these equivalent modifications or substitutions are all included within the scope defined by the claims of the present disclosure.

Claims

1. A method for controlling the anodizing process of aluminum alloys, characterized in that: include: Obtain the material grade and target film parameters of the workpiece to be processed, and generate initial process parameters; Based on the initial process parameters, a start-up reaction command is generated to initiate the anodizing reaction; Collect multi-dimensional process status data during the anodizing reaction to obtain multi-dimensional process status data; Multi-dimensional process status data are input into a preset long short-term memory neural network model to obtain deviation adjustment parameters; The process parameters during the anodizing reaction are adjusted according to the deviation adjustment parameters, and the anodizing reaction continues based on the adjusted process parameters. After a preset time period, the membrane parameters are acquired, and it is determined whether the membrane parameters have reached the preset target value. If so, a termination reaction command is generated.

2. The method for controlling the aluminum alloy anodizing process according to claim 1, characterized in that: The process of obtaining the material grade and target film parameters of the workpiece to be processed, and generating initial process parameters, specifically includes: Obtain the material grade and target film parameters of the workpiece to be processed; Based on the material grade and target membrane parameters, the corresponding electrolyte formula and pulse power parameters are retrieved from the preset material process parameter database; The obtained electrolyte formula and pulse power supply parameters are used as the initial process parameters.

3. The method for controlling the aluminum alloy anodizing process according to claim 1, characterized in that: The acquisition of multi-dimensional process states during the anodizing reaction to obtain multi-dimensional process state data specifically includes: Electrolyte state data are acquired at a first preset frequency; The membrane state data was acquired at a second preset frequency; Power output status data is collected at a third preset frequency; Electrolyte state data, membrane state data, and power output state data are integrated to obtain multi-dimensional process state data.

4. The method for controlling the aluminum alloy anodizing process according to claim 1, characterized in that: The step of inputting multi-dimensional process state data into a preset long short-term memory neural network model to obtain deviation adjustment parameters specifically includes: Multi-dimensional process status data are input into a pre-set long short-term memory neural network model; Temporal features are extracted from multi-dimensional process state data based on a long short-term memory neural network model to obtain temporal features; The timing characteristics are analyzed and processed to obtain the process state replenishment amount and process state adjustment value; The process condition replenishment amount and process condition adjustment value are integrated to obtain the deviation adjustment parameter.

5. The method for controlling the aluminum alloy anodizing process according to claim 1, characterized in that: The process parameters during the anodizing reaction are adjusted according to the deviation adjustment parameters, and the anodizing reaction continues based on the adjusted process parameters. Specifically, this includes: The deviation adjustment parameters are extracted to obtain the output waveform adjustment value, output power adjustment value, reaction parameter adjustment value, and component concentration replenishment amount; The output waveform of the pulse power supply is adjusted according to the output waveform adjustment value; The output power of the pulse power supply is adjusted according to the output power adjustment value; The concentration of the electrolyte components is adjusted according to the amount of component replenishment. The reaction parameters of the oxidation tank are adjusted according to the reaction parameter adjustment values; The anodic oxidation reaction continues based on the adjusted output waveform and power of the pulse power supply, the adjusted component concentration of the electrolyte, and the adjusted reaction parameters of the oxidation tank.

6. The method for controlling the aluminum alloy anodizing process according to claim 1, characterized in that: The process of acquiring membrane parameters after a preset time period and determining whether the membrane parameters have reached a preset target value, and if so, generating a termination reaction command, specifically includes: Obtain the current film thickness and current film impedance; Compare the current film thickness with the preset target film thickness range; Compare the current film impedance with a preset impedance threshold; If the current membrane thickness meets the preset target membrane thickness range and the membrane impedance detected three times consecutively is greater than or equal to the preset impedance threshold, then the membrane parameters are determined to have reached the preset target value, and a termination reaction command is generated.

7. The method for controlling the aluminum alloy anodizing process according to claim 1, characterized in that: Before acquiring multi-dimensional process status data during the anodizing reaction, the process further includes: Obtain historical process datasets; Construct an initial long short-term memory neural network model; The initial long short-term memory neural network model was trained using a historical process dataset to obtain a pre-defined long short-term memory neural network model.

8. A control system for anodizing aluminum alloy processes, characterized in that: The generation module is used to obtain the material grade and target film parameters of the workpiece to be processed, and to generate initial process parameters; The startup module is used to generate a startup reaction command based on the initial process parameters to initiate the anodizing reaction; The acquisition module is used to collect multi-dimensional process status during the anodizing reaction to obtain multi-dimensional process status data. The parameter generation module is used to input multi-dimensional process state data into a preset long short-term memory neural network model to obtain deviation adjustment parameters; The adjustment module is used to adjust the process parameters during the anodizing reaction according to the deviation adjustment parameters, and to continue the anodizing reaction based on the adjusted process parameters; The judgment module is used to acquire membrane parameters after a preset time period and determine whether the membrane parameters have reached the preset target value. If so, a termination reaction command is generated.

9. A control device for aluminum alloy anodizing process, characterized in that: The control device for the aluminum alloy anodizing process includes: a memory and at least one processor, wherein the memory stores instructions; at least one processor invokes the instructions in the memory to cause the control device for the aluminum alloy anodizing process to execute the various steps of the control method for the aluminum alloy anodizing process as described in any one of claims 1-7.

10. A computer-readable storage medium having instructions stored thereon, characterized in that, When the instructions are executed by the processor, they implement the various steps of the control method for the aluminum alloy anodizing process as described in any one of claims 1-7.

Images