Real-time monitoring and process parameter self-adaptive adjustment system for aluminum trim strip oxidation treatment

Abstract

The present application relates to the technical field of aluminum trim strip oxidation treatment, and discloses an aluminum trim strip oxidation treatment real-time monitoring and process parameter self-adaptive adjustment system, which comprises a process data acquisition module, a parameter optimization module, a surface monitoring module, a parameter matching module, a process adjustment module and a parameter self-learning module. The process data acquisition module acquires data in real time and generates relevant characteristic values and parameter adjustment sets; the parameter optimization module generates individualized process parameters and control strategies; the surface monitoring module acquires and screens abnormal surface data; the parameter matching module matches data to generate matching degrees and select target parameters; the process adjustment module drives correction parameters; and the parameter self-learning module updates a process database according to feedback data. The system realizes real-time monitoring and self-adaptive adjustment of process parameters during oxidation treatment, improves product quality and production efficiency, and is suitable for aluminum trim strip oxidation treatment.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of aluminum trim strip oxidation treatment, in particular to an aluminum trim strip oxidation treatment real-time monitoring and process parameter self-adaptive adjustment system. BACKGROUND

[0002] In the production and manufacturing process of aluminum trim strips, oxidation treatment is one of the key processes. The purpose is to form a uniform, dense and excellent performance oxide film on the surface of the aluminum trim strip to improve its corrosion resistance, wear resistance, decorative properties and other properties. However, the traditional aluminum trim strip oxidation treatment process has many problems, which is difficult to meet the requirements of modern industrial production for high quality, high efficiency and high stability.

[0003] From the aspect of process parameter control, the traditional process usually adopts fixed process parameters for oxidation treatment. However, in actual production, the process parameters of the oxidation tank (such as current density, oxidation liquid concentration, temperature, etc.) will be affected by various factors and fluctuate, such as changes in equipment operating state, differences in raw materials, different production batches, etc. Fixed process parameters cannot be adjusted according to these real-time changes, resulting in uneven quality of aluminum trim strip oxide films in different batches or even the same batch. For example, when the current density is unstable, it may cause uneven growth rate of the oxide film, thereby affecting the thickness and uniformity of the film layer; changes in oxidation liquid concentration will affect the oxidation reaction, and too high or too low concentration may cause oxide film quality defects, such as loose film layer, uneven color, etc.; fluctuations in temperature will affect the reaction kinetics, and thus affect the performance and appearance of the oxide film.

[0004] In terms of surface quality monitoring, the traditional method often relies on manual detection, which is not only inefficient, but also the detection results are easily affected by human factors, making it difficult to achieve real-time and accurate monitoring of the surface quality during the oxidation process of aluminum trim strips. Manual detection cannot timely detect abnormal conditions during the oxidation process, such as color difference, spots, scratches and other defects on the surface of the oxide film, which leads to the inability to timely adjust the process parameters, resulting in the production of a large number of unqualified products, increasing production costs and reducing production efficiency.

[0005] In addition, the traditional aluminum trim strip oxidation treatment process lacks self-adaptive adjustment capability. When process parameter fluctuations or surface quality abnormalities occur during production, it is difficult to quickly and accurately find the corresponding adjustment parameters, and operators need to rely on experience to make trial adjustments, which not only consumes time and effort, but also the adjustment effect is difficult to guarantee. The experience and level of different operators differ, resulting in poor consistency and stability of process adjustment, further affecting the quality and efficiency of aluminum trim strip oxidation treatment.

[0006] With the development of industrial automation and intelligence, higher requirements are put forward for the oxidation treatment process of aluminum trim. There is an urgent need for a system that can monitor the process parameters and surface quality in real time during the oxidation process and automatically adjust the process parameters according to real-time data, so as to realize precise control of aluminum trim oxidation treatment, improve the stability and consistency of product quality, reduce production cost, and improve production efficiency. The existing aluminum trim oxidation treatment technology has obvious deficiencies in real-time monitoring and adaptive adjustment, and cannot meet the needs of modern industrial production. Therefore, it is of great practical significance to develop a new type of aluminum trim oxidation treatment real-time monitoring and process parameter adaptive adjustment system. SUMMARY

[0007] The purpose of the present application is to provide an aluminum trim oxidation treatment real-time monitoring and process parameter adaptive adjustment system to solve the problems raised in the background art.

[0008] To achieve the above purpose, the present application provides the following technical solution: an aluminum trim oxidation treatment real-time monitoring and process parameter adaptive adjustment system, the system comprising:

[0009] A process data acquisition module is used to acquire process parameter data and surface state data of an aluminum trim oxidation tank in real time through an industrial Internet of Things terminal device, analyze and process to generate an oxidation film growth rate characteristic value and a tank liquid dynamic parameter characteristic value, and generate a process parameter dynamic adjustment set based on an initial process parameter set stored in a process database;

[0010] A parameter optimization module is used to process the individualized process parameters of the current batch of aluminum trim based on the oxidation film growth rate characteristic value and generate a dynamic oxidation control strategy;

[0011] A surface monitoring module is used to acquire real-time surface quality data during the oxidation process of the aluminum trim through a multi-spectral sensor, filter out abnormal surface data within the dynamic adjustment range and send it to the parameter matching module;

[0012] A parameter matching module is used to perform multi-dimensional matching of abnormal surface data and each parameter item in the process parameter dynamic adjustment set, generate a matching degree of real-time data and each process parameter item, and select the process parameter corresponding to the highest matching degree as the target adjustment parameter;

[0013] A process adjustment module is used to receive the target adjustment parameter and call a preset parameter adjustment protocol in the process database to drive the oxidation tank to perform a process parameter correction operation.

[0014] Preferably, the process of acquiring process parameter data of the aluminum trim oxidation tank in real time through the industrial Internet of Things terminal device is as follows:

[0015] identifying a unique identification code of the oxidation tank equipment, if it is a newly connected equipment, collecting basic process parameters including initial current density, oxidation liquid concentration reference value and temperature control range, establishing a process characteristic node based on the first data and performing parameter calibration, and generating an oxide film growth rate characteristic value;

[0016] If it is a historical access device, extract the historical process data set and surface quality change curve of the device, the historical process data set includes current fluctuation extreme value, concentration deviation record and temperature control delay time, and mark it as the initial parameter set of the current process analysis.

[0017] Preferably, the process parameter dynamic adjustment set is generated based on the initial process parameter set stored in the process database, specifically including:

[0018] extracting the initial standard set and dynamic correction coefficient set of each process parameter from the process database, the initial standard set including: current density safety range, concentration fluctuation tolerance interval and temperature regulation reference value;

[0019] The dynamic correction coefficient set includes current density compensation coefficient, concentration gradient adjustment coefficient and temperature response weight parameter;

[0020] Based on the real-time tank liquid dynamic parameter characteristic value, the initial standard set of each process parameter is dynamically adjusted, and the adjusted parameter set is recorded as the process parameter dynamic adjustment set;

[0021] The process parameter dynamic adjustment set includes the current dynamic range, concentration adaptability threshold and temperature optimization control value of each parameter item.

[0022] Preferably, based on the oxide film growth rate characteristic value, the individualized process parameters of the current batch of aluminum decorative strips are processed and a dynamic oxidation control strategy is generated, specifically including:

[0023] According to the mode matching of the oxide film growth rate characteristic value and the preset oxidation process characteristic library, the process parameter adjustment priority sequence is determined;

[0024] Based on the adjustment priority sequence, an adaptive control strategy containing parameter trigger conditions, adjustment step rules and fault handling mechanism is generated.

[0025] Preferably, the real-time surface quality data of the aluminum decorative strip during the oxidation process is obtained by a multi-spectral sensor, specifically including:

[0026] Monitoring the real-time spectral data stream of the oxide film formation process, including film thickness change value, surface uniformity index and color difference fluctuation value;

[0027] When the real-time spectral data stream exceeds the preset qualified process range, an abnormality mark is activated and surface data of an abnormal period is extracted as effective monitoring data.

[0028] Preferably, the matching degree of the generated real-time data and each process parameter item specifically includes:

[0029] The thickness deviation, uniformity change gradient, and color difference fluctuation amplitude in the abnormal surface data are analyzed, and difference value calculations are respectively performed on the current dynamic range, concentration adaptability threshold, and temperature optimization control value of each process parameter;

[0030] Based on the difference value calculation results, a matching degree index of the real-time data and each process parameter item is generated.

[0031] Preferably, the process parameter corresponding to the highest matching degree is selected as the target adjustment parameter, specifically including:

[0032] A matching degree sorting list of each process parameter item is established, and the parameter item corresponding to the first matching degree in the list is selected.

[0033] If the first matching degree is lower than a preset adjustment trigger threshold, a backup parameter set is called and matching degree calculation is performed again.

[0034] Preferably, based on the difference value calculation results, a matching degree index of the real-time data and each process parameter item is generated, and the specific processing process is:

[0035] A multi-dimensional matching algorithm is used to standardize the current difference value, concentration gradient difference, and temperature deviation, and generate a process parameter matching degree value ranging from 0 to 100;

[0036] The closer the matching degree value is to 100, the stronger the adaptability of the real-time data to the process parameter.

[0037] Preferably, the system further includes a parameter self-learning module, specifically including:

[0038] The oxidation film quality feedback data after each process adjustment is recorded, including actual film thickness uniformity, surface glossiness change, and color difference control effect;

[0039] The feedback data is reversely verified with the process parameter dynamic adjustment set, a parameter correction factor is generated, and is updated to the dynamic correction coefficient set of the process database.

[0040] Preferably, the parameter correction factor is generated, specifically including:

[0041] Based on the deviation degree of the feedback data from the expected process target, a current compensation factor, a concentration adjustment factor, and a temperature optimization weight are calculated;

[0042] The exponential weighted average algorithm is used to dynamically smooth the historical correction factors to generate a new set of dynamic correction coefficients.

[0043] Compared with the prior art, the present application has the following advantages:

[0044] In terms of real-time monitoring capability, the process data acquisition module can collect process parameter data and surface state data of the aluminum decorative strip oxidation tank in real time through the industrial Internet of Things terminal device. For newly connected devices, basic process parameters such as initial current density and oxidation liquid concentration reference value can be collected, process feature nodes can be established, and parameter calibration can be performed; for historical connected devices, historical process data sets and surface quality change curves are extracted, providing rich and accurate data support for subsequent process analysis and adjustment. The surface monitoring module uses a multi-spectral sensor to obtain real-time spectral data streams during the oxidation process of the aluminum decorative strip, including film thickness change values, surface uniformity indicators, and color difference fluctuation values, etc., which can timely detect abnormal surface quality during the oxidation process, achieving dynamic and real-time monitoring of the oxidation process, and changing the lag and inaccuracy of traditional manual detection.

[0045] In terms of adaptive adjustment of process parameters, the system can automatically generate a dynamic adjustment set of process parameters and individualized process parameters based on real-time monitoring data. The process data acquisition module dynamically adjusts the initial standard set of process parameters based on the initial process parameter set stored in the process database and the real-time bath liquid dynamic parameter characteristic values, generates a dynamic adjustment set of process parameters, including current dynamic range, concentration adaptability threshold, and temperature optimization control value, etc. The parameter optimization module determines the process parameter adjustment priority sequence by pattern matching the oxidation film growth rate characteristic value with the preset oxidation process feature library, generates an adaptive control strategy containing parameter trigger conditions, adjustment step rules, and fault handling mechanisms, and realizes individualized process parameter setting for the current batch of aluminum decorative strips. When the surface monitoring module detects abnormal surface data, the parameter matching module matches the abnormal surface data with each parameter item in the dynamic adjustment set of process parameters through a multi-dimensional matching algorithm, generates a matching degree index of real-time data and each process parameter item, selects the process parameter corresponding to the highest matching degree as the target adjustment parameter, and the process adjustment module receives the target adjustment parameter and drives the oxidation tank to perform process parameter correction operations, realizing rapid and accurate adjustment of process parameters, improving the efficiency and accuracy of process adjustment, and avoiding the blindness and uncertainty of traditional manual experience adjustment.

[0046] In terms of system self-learning and optimization, the parameter self-learning module records the oxide film quality feedback data after each process adjustment, including actual film thickness uniformity, surface glossiness change and color difference control effect, etc., and performs reverse verification on the feedback data and the process parameter dynamic adjustment set to generate a parameter correction factor. The historical correction factor is dynamically smoothed by an exponential weighted average algorithm, and the dynamic correction coefficient set of the process database is updated, so that the system can continuously accumulate experience, automatically optimize the process parameters according to the actual production situation, improve the adaptability and stability of the system, and the control effect of the system will be continuously improved as the production proceeds, further ensuring the stability and consistency of the aluminum trim strip oxidation treatment quality.

[0047] Through the synergistic effect of the above aspects, the system realizes automatic and intelligent control of the aluminum trim strip oxidation treatment process, significantly improves the product quality, reduces the unqualified rate, reduces the cost of manual detection and process adjustment, improves the production efficiency, has good economic and social benefits, and provides a new direction and idea for the development of aluminum trim strip oxidation treatment technology. BRIEF DESCRIPTION OF DRAWINGS

[0048] Fig. 1 The working principle diagram of the aluminum trim strip oxidation treatment real-time monitoring and process parameter self-adaptive adjustment system described in the application is shown in the figure.

[0049] Fig. 2 The working principle diagram of the process data acquisition module is shown in the figure.

[0050] Fig. 3 The principle diagram for generating the process parameter dynamic adjustment set is shown in the figure. DETAILED DESCRIPTION

[0051] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, not all. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the application.

[0052] Please refer to Figs. 1-3 The aluminum trim strip oxidation treatment real-time monitoring and process parameter self-adaptive adjustment system described in the application comprises:

[0053] A process data acquisition module, a parameter optimization module, a surface monitoring module, a parameter matching module and a process adjustment module.

[0054] The process data acquisition module performs the following operations through the industrial Internet of Things terminal device: real-time acquisition of process parameter data (such as current density, oxidation liquid concentration, temperature, etc.) and surface state data of the aluminum decorative strip oxidation tank, analysis and processing of the collected data, generation of oxidation film growth rate characteristic values and tank liquid dynamic parameter characteristic values (such as concentration change gradient, temperature response delay, etc.); at the same time, based on the initial process parameter set (including current density safety range, concentration fluctuation tolerance interval, etc. basic parameters) stored in the process database, combined with real-time data to generate a process parameter dynamic adjustment set (containing current dynamic range, concentration adaptability threshold, etc. adjustable parameter items).

[0055] The parameter optimization module determines the process parameter adjustment priority sequence (such as current density adjustment priority higher than temperature adjustment) based on the oxidation film growth rate characteristic value, and matches it with the preset oxidation process characteristic library, and generates a dynamic oxidation control strategy containing parameter trigger conditions (such as triggering current adjustment when the film thickness growth rate is lower than the threshold), adjustment step rules (such as increasing the current density in stages), and fault handling mechanisms (such as out-of-range alarm shutdown), and accordingly obtains the individualized process parameters of the current batch of aluminum decorative strips.

[0056] The surface monitoring module continuously acquires real-time spectral data streams during the oxidation process of the aluminum decorative strip through the multispectral sensor, including film thickness change value, surface uniformity index, and color difference fluctuation value, etc. Real-time surface quality data; when it is monitored that the real-time spectral data stream exceeds the preset qualified process range, activate the abnormal marker and extract the surface data (such as thickness deviation, uniformity mutation value, etc.) of the abnormal period as effective monitoring data, and send the abnormal surface data that meet the dynamic adjustment range to the parameter matching module.

[0057] The parameter matching module receives the abnormal surface data, analyzes the thickness deviation, uniformity change gradient, and color difference fluctuation amplitude, etc. characteristic parameters, and performs difference calculation with the current dynamic range, concentration adaptability threshold, and temperature optimization control value in the process parameter dynamic adjustment set; using a multi-dimensional matching algorithm to standardize the current difference, concentration gradient difference, and temperature deviation, generate a process parameter matching degree value (the closer the matching degree value is to 100, the stronger the adaptability) ranging from 0 to 100, establish a matching degree sorting list of each process parameter item, and select the parameter item corresponding to the first matching degree in the list as the target adjustment parameter; if the first matching degree is lower than the preset adjustment trigger threshold (such as lower than 60), a backup parameter set is called and matching degree calculation is performed again.

[0058] The process adjustment module receives target adjustment parameters, calls preset parameter adjustment protocols (such as current adjustment step rules, concentration addition ratio rules, etc.) in the process database, drives the oxidation tank to perform process parameter correction operations (such as automatically adjusting the power supply output current, starting the liquid supplement pump to add concentrated liquid, etc.), and realizes real-time correction of process parameters.

[0059] The application will be further described below in combination with examples 1 to 5:

[0060] Example 1

[0061] The process data acquisition module of the system performs differentiated data acquisition processes according to the access state of the oxidation tank equipment. For newly connected equipment, the equipment identity is first identified through the equipment unique identification code, which can be obtained through the equipment nameplate identification, the preset equipment ID field in the industrial Internet of Things protocol, or the electronic tag built-in the sensor module, etc. After identifying the newly connected equipment, the system starts the initial data acquisition program to collect basic process parameters, including the initial current density, the oxidation liquid concentration reference value and the temperature control range. Among them, the collection range of the initial current density is preset according to the safety interval of the oxidation process type and the aluminum decoration strip material, for example, for the conventional aluminum alloy decoration strip, the initial current density can be collected as 1.5-2.5 A / dm², which is determined by the industry standard process parameters stored in the process database; the collection of the oxidation liquid concentration reference value is based on the type of oxidation liquid, for example, when using sulfuric acid anodic oxidation process, the concentration reference value is collected as 180-200 g / L, which meets the conventional concentration requirement of sulfuric acid anodic oxidation process; the temperature control range is set according to the thermodynamic characteristics of the oxidation reaction, usually collected as 20-30℃, to ensure the stability of the oxidation film growth.

[0062] Based on the first collected basic process parameter data, the system establishes the process feature node. The establishment process of the process feature node includes the association and mapping of parameters such as initial current density, concentration reference value and temperature control range with the oxidation film growth rate, for example, through the measured value of the film thickness growth rate at the initial current density of 1.5 A / dm², the concentration of 180 g / L and the temperature of 20℃, the coordinate parameters of the first process feature node are established. Then the parameter calibration operation is performed, and the calibration process is realized by comparing the deviation between the theoretical calculation value and the measured value. The theoretical calculation value is obtained based on Faraday's law and the oxidation film growth model, for example, the theoretical film thickness is calculated according to the current density and time, and compared with the measured film thickness. If the deviation exceeds the preset threshold (such as ±5%), the initial parameters are corrected, and the correction methods include linear interpolation adjustment or proportion coefficient correction, and finally the oxidation film growth rate characteristic value is generated, which includes the film thickness growth amount per unit time, the growth rate fluctuation coefficient and other parameters.

[0063] For historical access equipment, the system retrieves the process database through the equipment unique identification code, extracts the historical process data set and surface quality change curve of the equipment. The historical process data set contains historical parameters such as current fluctuation extreme value, concentration deviation record and temperature control delay time length. The current fluctuation extreme value records the maximum and minimum values of the current density of the equipment in the past oxidation process, for example, the historical current fluctuation extreme value of a certain equipment is 1.2-2.8 A / dm², which reflects the current stability of the equipment in different batches of production; the concentration deviation record stores the amplitude and duration of the concentration deviation from the reference value, such as a certain concentration deviation of +15 g / L for 30 minutes, which reflects the fluctuation of concentration control; the temperature control delay time length refers to the time interval from the system receiving the temperature adjustment instruction to the actual temperature reaching the target value, for example, the temperature control delay time length of a certain equipment is 15 minutes, which reflects the response characteristics of the equipment temperature control system.

[0064] The surface quality change curve is generated by the surface monitoring data in the historical production process, including film thickness uniformity change trend, surface gloss fluctuation curve and color difference evolution trajectory, etc. For example, in the oxidation process of a certain batch of aluminum decorative strip, the film thickness uniformity standard deviation gradually increases from 2 μm to 5 μm within 0-30 minutes, and then decreases to 3 μm after adjusting the current density, forming a specific fluctuation curve; the surface gloss increases rapidly at the beginning of oxidation, reaches a peak value and then tends to be stable, reflecting the stage characteristics of the oxide film formation; the color difference fluctuation curve records the degree of color deviation from the standard value during the oxidation process, such as the ΔE value in CIELab color space gradually increases from 1.0 to 2.5, and then decreases to 1.5 through process adjustment.

[0065] The system marks the extracted historical process data set and surface quality change curve as the initial parameter set of the current process analysis. The initial parameter set is used as the reference for the current batch production, which is used for trend analysis and deviation comparison when dynamically adjusting the process parameters. For example, in the current batch production, if the real-time monitored current density fluctuation approaches the historical fluctuation extreme value, the system can trigger an early warning mechanism or adjust the dynamic correction coefficient to avoid exceeding the safety range; if the surface quality change curve shows that the current film thickness uniformity change trend is similar to that of a certain batch in history, the system can directly call the process parameter adjustment strategy of that batch to improve the adjustment efficiency.

[0066] During data collection, the industrial IoT terminal device acquires multi-dimensional data through a sensor network. The current density data is collected in real time by a Hall current sensor, which is installed in the power circuit of the oxidation tank and has an accuracy of ±1%; the concentration of the oxidation solution is measured by an online concentration meter, such as a conductivity sensor or a refractometer, which converts the concentration value according to the solution conductivity or refractive index, with a resolution of 0.1 g / L; the temperature data are collected by a Pt100 temperature sensor, which is arranged in the solution of the oxidation tank and has a temperature measurement accuracy of ±0.5°C. The surface state data are acquired by image acquisition equipment or spectral sensors, such as a line array CCD camera for shooting the surface image of the aluminum trim strip, or a multi-spectral sensor for collecting reflectance spectrum data, which are used for subsequent analysis of the growth rate of the oxidation film and the surface quality.

[0067] In terms of data transmission, the industrial IoT terminal device transmits the collected real-time data to the processing unit of the process data collection module through wired networks (such as Ethernet) or wireless networks (such as Wi-Fi, 5G). The processing unit adopts an edge computing architecture to preprocess the raw data, including denoising filtering, data normalization, and feature extraction. Denoising filtering removes random noise through median filtering or Kalman filtering algorithms to ensure data reliability; data normalization converts parameters of different dimensions to values within a uniform range, facilitating subsequent analysis; feature extraction extracts the oxidation film growth rate characteristic value and the tank liquid dynamic parameter characteristic value from the raw data, such as the concentration change gradient obtained by differential calculation and the temperature response delay duration obtained by delay estimation algorithm.

[0068] The process data collection module also has data storage and query functions, storing the collected historical data and generated characteristic values in the process database to form device archives and process history records. The stored data can be used for subsequent process optimization, device performance evaluation, and fault tracing. For example, by analyzing the historical process data set of a device, it can be found that the temperature control delay duration increases with the increase of the service life of the device, so that a targeted maintenance plan can be developed; by comparing the surface quality change curves of different batches, the initial process parameter set can be optimized to improve process stability.

[0069] Example 2:

[0070] When the process data collection module of the system generates the dynamic adjustment set of process parameters, it needs to extract the initial standard set and the dynamic correction coefficient set from the process database, and combine the real-time tank liquid dynamic parameter characteristic value to adjust the parameters. The following details this process in combination with specific scenarios:

[0071] Suppose an aluminum trim strip oxidation tank uses sulfuric acid anodic oxidation process, the initial standard set stored in the process database includes: the current density safety range is 1.0-3.0 A / dm2, the concentration fluctuation tolerance interval is ±10 g / L (based on the reference value of 200 g / L), and the temperature regulation reference value is 25°C. In the dynamic correction coefficient set, the current density compensation coefficient is ±0.1 A / dm2 / μm (indicating that for every 1 μm deviation from the target film thickness growth rate, the current density is adjusted by 0.1 A / dm2), the concentration gradient adjustment coefficient is ±5 g / L / ‰ (for every 1 ‰ / min detected concentration change rate, the concentration threshold is adjusted by 5 g / L), and the temperature response weight parameter is 0.8-1.2 (used to adjust the influence weight of temperature on film thickness growth rate).

[0072] In a certain batch production, the real-time tank liquid dynamic parameter characteristic values are collected by online sensors as follows: the oxidation liquid concentration measured value is 185 g / L, and it is continuously decreasing at a rate of 2 ‰ / min; the temperature measured value is 28°C, with a fluctuation amplitude of ±1.5°C; the current film thickness growth rate is 0.18 μm / min, which is lower than the preset target value 0.2 μm / min. Based on these real-time data, the system starts the dynamic adjustment process:

[0073] First, process the concentration parameter. The initial concentration reference value is 200 g / L, and the measured value 185 g / L has deviated from the reference value by 15 g / L, which exceeds the initial tolerance interval ±10 g / L. At the same time, the concentration change rate is -2 ‰ / min (negative sign indicating decrease), which exceeds the trigger threshold 1 ‰ / min. According to the concentration gradient adjustment coefficient ±5 g / L / ‰ in the dynamic correction coefficient set, the concentration adaptive threshold adjustment amount is calculated as: 2 ‰ / min × 5 g / L / ‰ = 10 g / L. Therefore, the lower limit of the concentration dynamic adjustment is adjusted from the initial 190 g / L (200-10) to 180 g / L (200-10-10), and the upper limit remains unchanged at 210 g / L, forming a new concentration adaptive threshold of 180-210 g / L.

[0074] Then adjust the current density parameter. Since the film thickness growth rate is lower than the target value 0.02 pm / min, according to the current density compensation coefficient ±0.1 A / dm2 / pm, the current density compensation amount is calculated as -0.02 pm / min x 0.1 A / dm2 / pm = -0.002 A / dm2(the negative sign indicates that the current density needs to be reduced to slow down the growth rate). But need to be combined with the current density safety range judgment: the initial safety range is 1.0-3.0 A / dm2, the current measured current density is 2.5 A / dm2, and after adjustment it is 2.5-0.002 = 2.498 A / dm2, still within the safety range. Therefore, the current dynamic range keeps the initial lower limit 1.0 A / dm2 unchanged, and the upper limit is adjusted to 2.5 A / dm2 (based on the current measured value and compensation amount, to avoid excessive adjustment), forming a new current dynamic range of 1.0-2.5 A / dm2.

[0075] In terms of temperature parameter adjustment, the measured temperature is 28°C, which is higher than the reference value 25°C, and the fluctuation amplitude exceeds the initial allowed range (usually default fluctuation amplitude ±1°C). The temperature response weight parameter is used to adjust the influence weight of temperature on the film thickness growth rate. The current measured temperature is too high, so the system automatically increases the temperature response weight from the default value 1.0 to 1.1 to enhance the adjustment effect of temperature on the growth rate. Based on the temperature regulation reference value 25°C, combined with the real-time temperature 28°C, the temperature optimization control value is calculated as 25°C+(28°C-25°C) x 1.1 = 28.3°C, but it needs to be limited within the temperature range allowed by the process (such as 20-30°C), so the final temperature optimization control value is set to 28°C, and the temperature control priority is raised to the second place in the current adjustment sequence (only next to the concentration adjustment).

[0076] In the above adjustment process, the system needs to verify the relevance of each parameter adjustment. For example, the decrease in concentration may lead to a decrease in the conductivity of the solution, and if the current density is also reduced, it may further affect the uniformity of the film thickness growth. Therefore, when generating the process parameter dynamic adjustment set, the system uses the parameter correlation model (such as the synergistic effect curve of concentration and current density) stored in the process database to ensure that the adjusted parameter combination (concentration 180-210 g / L, current 1.0-2.5 A / dm2, temperature 28°C) exists in the historical process data, avoiding parameter conflicts.

[0077] In another scenario, if the impurity ion concentration in the oxidation tank solution increases (such as an increase in copper ion content) due to long-term use, real-time monitoring finds that the characteristic value of the dynamic parameter of the solution appears abnormal: the concentration change rate fluctuates sharply (±3‰ / min), and the temperature control delay time length is extended from the historical average of 10 minutes to 15 minutes. At this time, the system automatically retrieves the standby set of dynamic correction coefficients from the dynamic correction coefficient set: the concentration gradient adjustment coefficient is temporarily adjusted from ±5 g / L / ‰ to ±8 g / L / ‰ to cope with the rapid concentration fluctuation; the lower limit of the temperature response weight parameter is reduced from 0.8 to 0.6 to reduce the sensitivity of temperature adjustment and avoid over-adjustment due to temperature control delay. After adjustment, the concentration adaptive threshold range is expanded to 175-225 g / L (200±25 g / L), the current dynamic range remains unchanged at 1.0-3.0 A / dm² (since the impurity effect is mainly related to concentration), and the temperature optimization control value is set to 25℃±2℃ (the fluctuation tolerance interval is expanded).

[0078] In terms of real-time parameter adjustment, the system realizes one-second data acquisition and analysis through the edge computing unit of the industrial Internet of Things terminal device, ensuring that the dynamic adjustment set is updated every 5 minutes. For example, when the concentration change rate slows down from -2‰ / min to -1‰ / min within 10 minutes, the system automatically adjusts the lower limit of the concentration adaptive threshold from 180 g / L to 185 g / L at the second update, gradually converging to the vicinity of the initial reference value, reflecting the gradual nature of adaptive adjustment.

[0079] After generating the process parameter dynamic adjustment set, the system synchronizes it to the parameter optimization module and the parameter matching module. Based on the current dynamic range and concentration adaptive threshold in the set, combined with the material characteristics of the current batch of aluminum decorative strips (such as the lower oxidation difficulty of 6063 aluminum alloy), the parameter optimization module further refines the individualized parameters: the current density initial value is set to 2.0 A / dm² (at the median of the dynamic range), and the concentration control target value is set to 195 g / L (close to the reference value and leaving room for downward adjustment). The parameter matching module uses the set as a matching reference, which can quickly match with the current dynamic range and temperature optimization control value when the surface monitoring module detects a decrease in film thickness uniformity, determining the adjustment direction.

[0080] During the entire generation process, the process database serves as the core support, storing more than 1000 historical process parameter combinations and their corresponding surface quality data, forming an experience knowledge base for parameter adjustment. For example, when the system detects that concentration decreases and temperature increases simultaneously, it can retrieve historical adjustment cases under similar working conditions (such as adjusting the concentration lower limit to 180 g / L and the temperature upper limit to 28℃, which improves film thickness uniformity) from the database, directly calling the parameter adjustment amplitude of the case as a reference, shortening the calculation time and improving the adjustment accuracy.

[0081] Example 3:

[0082] The process of generating a dynamic oxidation control strategy based on the characteristic value of the oxide film growth rate by the parameter optimization module needs to be combined with the process feature library matching and parameter priority logic. The specific implementation is described in detail below:

[0083] Suppose a batch of aluminum decorative strips uses AA6061 aluminum alloy material, and the oxidation process target is to generate an anodic oxide film with a thickness of 20 μm. The preset oxidation process feature library stores film growth rate models under different combinations of current density, concentration, and temperature. The construction of the process feature library is based on historical process data, such as including typical parameter combinations such as working condition A (current density 2.0 A / dm², concentration 200 g / L, temperature 25 ℃, film growth rate 0.2 μm / min), working condition B (current density 2.5 A / dm², concentration 190 g / L, temperature 28 ℃, film growth rate 0.25 μm / min), and their corresponding growth rate characteristic values.

[0084] In real-time processing, the oxide film growth rate characteristic value generated by the process data acquisition module is 0.16 μm / min (measured value), which is lower than the target value 0.2 μm / min. The parameter optimization module first matches the characteristic value with the model in the process feature library, and uses the Euclidean distance algorithm to calculate the difference degree of the measured value and each working condition, the formula is:

[0085]

[0086] wherein, is the difference degree, is the measured growth rate (unit: μm / min), is the growth rate of the th working condition in the process feature library, is the measured concentration (unit: g / L), is the concentration of the corresponding working condition, is the measured temperature (unit: ℃), is the temperature of the corresponding working condition, and are the weight coefficients of concentration and temperature (default , , which indicates that the concentration has a more significant impact on the growth rate).

[0087] Through calculation, the difference degree of the measured value and working condition A is , and the difference degree of the measured value and working condition B is , the system determines that the process parameter adjustment priority sequence is: current density adjustment > concentration adjustment > temperature adjustment. The priority setting is based on the principle of oxidation process: current density directly affects the electrochemical reaction rate and is the most direct control factor for film thickness growth; concentration fluctuation affects solution conductivity, followed by temperature, which mainly affects the reaction activation energy, and has the lowest adjustment priority.

[0088] Based on the priority sequence, the system generates a dynamic oxidation control strategy. First, set the parameter trigger condition: when the growth rate is continuously lower than 0.18 pm / min for 10 minutes, trigger current density adjustment. The adjustment step rule is divided into three stages: in the first stage, the current density is increased from 1.8 A / dm² to 2.0 A / dm², and the monitoring is continued for 15 minutes; if the growth rate is still not up to standard (still <0.2 pm / min), in the second stage, the concentration is increased from 195 g / L to 200 g / L while maintaining the current density at 2.0 A / dm², and the monitoring is continued for another 15 minutes; if it is still not up to standard, in the third stage, the temperature is increased from 23°C to 25°C, forming a combined adjustment.

[0089] The fault handling mechanism includes: if the growth rate is still lower than 0.15 pm / min after three adjustments, it is determined to be an abnormal condition, triggering an alarm and automatically suspending the operation of the oxidation tank, while generating a fault log (recording the current parameters, adjustment history and abnormal characteristics) for manual troubleshooting. In addition, in the personalized parameter processing of AA6061 material, the system retrieves the historical correction coefficient of this material (such as current density sensitivity coefficient +0.1 A / dm²·pm - ¹) from the process database, and reduces the current density adjustment step from the default 0.2 A / dm² to 0.1 A / dm², to avoid excessive current impact due to differences in material conductivity.

[0090] In another application scenario, if the measured oxidation film growth rate is 0.25 pm / min (higher than the target value 0.2 pm / min), the system matches the process feature library and finds that it matches the condition B (high current density, low concentration, high temperature) mode. At this time, the adjustment priority sequence is: current density reduction > temperature reduction > concentration increase. The trigger condition is set to growth rate > 0.23 pm / min for 5 minutes, and the adjustment step rule is: first reduce the current density from 2.8 A / dm² to 2.5 A / dm², and at the same time reduce the temperature from 28°C to 25°C, forming a synergistic adjustment; if the growth rate does not fall back to the target interval after 30 minutes, increase the concentration from 185 g / L to 195 g / L to indirectly reduce the effective current density by increasing the solution resistance.

[0091] The parameter optimization module also needs to deal with the coupling effect of multiple parameters. For example, when the current density increases, the exothermic reaction of oxidation may cause the temperature to rise passively. The system can foresee this coupling relationship through historical data in the process database, and adjust the temperature control target value by 1°C in advance to offset the influence of reaction heat, so as to maintain the stability of the temperature. In addition, for the surface pretreatment differences of different batches of aluminum decorative strips (such as the influence of mechanical polishing degree on film formation rate), the system automatically adjusts the matching weight of the process feature library through the input pretreatment process code (such as P01 representing rough polishing and P02 representing fine polishing). The film formation rate of fine polishing workpiece is usually 10% faster than that of rough polishing, so the tolerance of growth rate deviation is expanded to ±0.02 μm / min during matching to avoid misadjustment caused by pretreatment differences.

[0092] After the dynamic oxidation control strategy is generated, the system converts it into an executable parameter instruction sequence, for example:

[0093] Current density adjustment instruction: +0.2 A / dm² (relative to the current value), execution time t=0 min;

[0094] Concentration monitoring instruction: detect the concentration value every 5 minutes, and trigger the liquid supplement pump if it is <190 g / L;

[0095] Temperature adjustment instruction: if the current density is ≥2.0 A / dm² after 30 min, start the cooling system to control the temperature at 24±1°C.

[0096] These instructions are transmitted to the control system of the oxidation tank through an industrial bus (such as PROFINET) to realize automatic adjustment of parameters. At the same time, the time parameters (such as monitoring interval, adjustment delay) in the strategy are all set based on the response characteristics of the equipment, for example, the current adjustment response time of the oxidation tank is 2 minutes, so the interval between two adjustment operations is at least 5 minutes to ensure that the effect of the previous adjustment is fully manifested.

[0097] The update mechanism of the process feature library runs throughout the entire process. After each batch production is completed, the system stores the actual executed process parameters (such as the final current density 2.2 A / dm², the concentration 198 g / L, and the temperature 25°C) and the corresponding growth rate (0.21 μm / min) into the process feature library to expand the coverage of the model. For newly emerging parameter combinations (such as current density 3.0 A / dm² and concentration 210 g / L), the system automatically marks them as “to be verified working conditions” and prioritizes the collection of surface quality data under this working condition in subsequent production to gradually improve the accuracy of the model.

[0098] Example 4:

[0099] The surface monitoring module obtains real-time surface quality data of the oxidation process of aluminum trim through a multispectral sensor. Taking the oxidation production of AA6063 aluminum alloy trim as an example, the process is described. The system uses a multispectral sensor based on visible-near infrared spectrum. The sensor integrates a spectral acquisition unit with a wavelength range of 400-1000 nm and is installed 20 cm above the outlet of the oxidation tank. It performs dynamic scanning of the aluminum trim surface on the conveyor belt at a vertical viewing angle, acquiring 10 sets of spectral data per second and recording the acquisition timestamp synchronously.

[0100] During the initial stage of oxide film formation (0-15 minutes), the sensor continuously monitors the spectral data stream and analyzes key parameters. The film thickness change value is calculated based on the spectral reflectance model. For example, when the aluminum substrate surface starts to form an oxide film, the reflectivity at 450 nm gradually decreases from 85% for aluminum. According to the mapping relationship between film thickness and reflectivity based on the Fresnel reflection formula, the real-time film thickness can be calculated. Assuming that the spectral data at a certain time shows a 450 nm reflectivity of 72%, corresponding to a film thickness of 2.3 μm, compared to the previous time (2.1 μm), the thickness change value is +0.2 μm. The surface uniformity index is obtained by calculating the spectral difference coefficient of different pixels in the same scanning line. For example, the trim surface is divided into 10 detection regions, 50 spectral samples are extracted from each region, and the standard deviation of the spectral mean of each region is calculated. If the standard deviation of a certain region is 8%, the uniformity index is 8%. The color difference fluctuation value is calculated using the CIELab color space. Taking the L*=85, a*=0, b*=5 of the aluminum material before oxidation as the reference value, when the real-time spectral data is converted to L*=83, a*=+1.2, b*=6.5, the color difference ΔE=√[(83-85)²+(1.2-0)²+(6.5-5)²]≈2.5.

[0101] When a batch of aluminum trim is oxidized to 20 minutes, the sensor detects the following abnormalities: the film thickness change value decreases from 0.15 μm / min to 0.08 μm / min within 18-20 minutes, the surface uniformity index increases from 5% to 12%, and the color difference fluctuation value ΔE increases from 1.8 to 3.0. These parameters exceed the pre-set qualified process range (thickness change rate ≥0.1 μm / min and ≤0.2 μm / min, uniformity ≤8%, ΔE ≤2.5), and the system immediately activates the abnormal marker and extracts the surface data of the abnormal period (18-22 minutes) as effective monitoring data. Specifically, it includes:

[0102] Continuous spectral sampling points: 10 sets of spectral data per second, a total of 24 sets, including wavelength reflectivity, film thickness calculation value, uniformity real-time value, and color difference ΔE value;

[0103] Corresponding process parameter timestamps: e.g. current density 2.2 A / dm2, concentration 195 g / L, temperature 26°C at 18:05, current density 2.3 A / dm2, concentration 193 g / L, temperature 27°C at 18:10, etc.

[0104] Surface image features: the pseudo-color image reconstructed from the spectral data shows a band-shaped low reflectivity area in the middle of the trim strip, which is presumably due to uneven thickness of the oxide film.

[0105] During the screening of abnormal surface data, the system first excludes accidental fluctuations caused by the vibration of the conveying belt (e.g. data with a vibration amplitude > 0.5 mm are automatically marked as invalid), and avoids false positives by setting a time window (e.g. only when 3 consecutive samplings exceed the threshold does an abnormality trigger). In this example, the thickness change rate is below 0.1 pm / min for 4 consecutive samplings, and the uniformity exceeds 10% for 2 consecutive samplings, meeting the abnormality triggering conditions, so the data is confirmed to be valid and sent to the parameter matching module.

[0106] The calibration mechanism of the multispectral sensor ensures data reliability. Before starting up each day, the system automatically calibrates the sensor with a whiteboard (reflectivity 100%) and dark current (blocking the light source to detect noise), and the spectral acquisition error is ≤ ± 2% after calibration. During production, a standard oxide film sample plate (thickness 10 pm, uniformity 3%, ΔE = 1.0) is inserted every hour for real-time verification. If the detected value deviates from the standard value by > 5%, the drift correction algorithm is automatically started, and the spectral-parameter conversion model is adjusted through polynomial fitting.

[0107] The monitoring focus is dynamically adjusted at different process stages. In the early stage of oxidation (0-10 minutes), the system focuses on monitoring whether the film thickness growth rate reaches the theoretical value (e.g. 0.15-0.2 pm / min), and if it is below the lower limit, the current density adjustment is triggered in advance; in the middle and late stages (10-30 minutes), uniformity and color difference become the main monitoring indicators, for example, when the film thickness approaches the target value of 20 pm, the uniformity needs to be controlled at ≤ 6%, and ΔE ≤ 2.0, otherwise the surface quality is improved by fine-tuning the concentration or temperature.

[0108] For aluminum trim strips with complex geometries (e.g. with grooves or curved structures), the multispectral sensor is equipped with a two-dimensional scanning mechanism to monitor different curvature surfaces in different zones. For example, the spectral collection angle of the trim strip groove area is adjusted to 45° to avoid direct light reflection interference, and a curvature compensation coefficient is added to the algorithm (e.g. when the groove depth > 2 mm, the film thickness calculation value is multiplied by a correction factor of 1.1), to ensure the consistency of data in different regions.

[0109] In terms of data transmission, real-time spectral data streams are transmitted to the edge computing server through Gigabit Ethernet. The FPGA chip on the server realizes real-time analysis of spectral data, with a delay of <500 ms from raw spectral data to the generation of parameters such as thickness, uniformity, and color difference. When abnormal data triggers, the edge computing server sends a warning signal to the parameter matching module in synchronization, starting multi-threaded processing: one continues to collect real-time data, and the other packages and stores abnormal period data to a solid state disk for subsequent traceability analysis.

[0110] The historical data storage strategy of the surface monitoring module adopts hierarchical storage: real-time data are retained for the last 24 hours for real-time analysis; daily data are stored on a disk array after compression, with a retention period of 3 months; key abnormal data (such as uniformity > 15% or ΔE > 3.0) are automatically backed up to a cloud server for long-term storage. Through historical data queries, the impact of different process parameter combinations on surface quality can be analyzed, for example, it is found that when the temperature is > 28°C and the concentration is < 190 g / L, the probability of uniformity exceeding the standard increases by 40%, providing data support for process optimization.

[0111] Example 5:

[0112] The parameter self-learning module of the system optimizes the process parameter set in reverse by recording the quality feedback data after process adjustment, realizing the adaptive evolution of the system. The working mechanism of this module is described in detail below in combination with the specific implementation process:

[0113] Suppose that during the oxidation process of a batch of aluminum decorative strips, the process adjustment module adjusts the current density from 2.0 A / dm² to 2.2 A / dm² based on the parameter matching results to improve the film thickness growth rate. After the adjustment is completed, the parameter self-learning module starts the data recording process and collects the following oxidation film quality feedback data:

[0114] Actual film thickness uniformity: The film thickness standard deviation σ value is calculated to be 4.2 μm (5.0 μm before adjustment) by scanning the full surface with a multi-spectral sensor;

[0115] Surface glossiness change: The average glossiness value is increased from 85 GU to 92 GU using a 60° glossiness meter;

[0116] Color difference control effect: In the CIELab color space, the final color difference ΔE value is reduced from 2.8 to 2.3, close to the target value of 2.0.

[0117] These feedback data are verified in reverse with the initial parameters in the dynamic adjustment set of process parameters (current dynamic range 1.8-2.5 A / dm², concentration adaptability threshold 190-210 g / L). The verification process includes:

[0118] Parameter correlation analysis: confirm the correlation between current density adjustment and film thickness uniformity improvement. Through historical data comparison, it is found that when the current density is in the range of 2.1-2.3 A / dm², the uniformity standard deviation is reduced by an average of 15%-20%, and the current adjustment conforms to this trend;

[0119] Target deviation calculation: the film thickness uniformity target value is σ≤4.0 μm, and the actual measured value is 4.2 μm, which still deviates from the target by 0.2 μm. The color difference target value ΔE≤2.0, and the actual measured value deviates by 0.3 μm, indicating that the adjustment effect does not fully meet the standard, and a correction factor needs to be generated.

[0120] Based on the deviation degree of feedback data and expected target, the system calculates the parameter correction factor. For current density, since the uniformity is improved but not yet reached the target, the current compensation factor is calculated as: (target σ - measured σ) x 0.1 A / dm² / μm = (4.0-4.2) x 0.1 = -0.02 A / dm² (negative sign indicates that the current density needs to be slightly reduced to further optimize the uniformity); for color difference, concentration adjustment has a more significant impact on color difference, and the concentration adjustment factor is calculated as: (measured ΔE - target ΔE) x 5 g / L / ΔE = (2.3-2.0) x 5 = +1.5 g / L (positive sign indicates that the concentration needs to be increased to reduce the color difference); for temperature optimization weight, the temperature remains at 25°C in this adjustment, and since the glossiness improvement is not directly related to temperature, the weight parameter remains at the default value of 1.0.

[0121] The generation of the correction factor needs to be combined with historical data smoothing. The system retrieves the correction factors of the previous three similar process adjustments: the first time, the current compensation factor is -0.05 A / dm², and the concentration adjustment factor is +2.0 g / L; the second time, -0.03 A / dm², +1.0 g / L, and the third time, -0.04 A / dm², +1.5 g / L. Using the exponential weighted average algorithm, the latest correction factor (-0.02 A / dm², +1.5 g / L) is dynamically smoothed with the historical values, giving the most recent correction factor a weight of 40%, the previous two a weight of 30% and 20% respectively, and the new current density compensation coefficient is calculated as: (-0.02 x 40%) + (-0.04 x 30%) + (-0.03 x 20%) = -0.03 A / dm² / μm

[0122] The concentration gradient adjustment coefficient is: (+1.5 x 40%) + (+1.5 x 30%) + (+1.0 x 20%) = +1.35 g / L / ΔE

[0123] The newly generated dynamic correction coefficient set is updated to the process database, replacing the original coefficients (original current compensation coefficient -0.01 A / dm² / μm, concentration adjustment coefficient +1.0 g / L / ΔE). In subsequent batch production, when similar conditions occur again (such as film thickness growth rate meeting the standard but uniformity being insufficient), the system will preferentially use the updated coefficients for parameter adjustment, for example, adjusting the current density adjustment step from 0.2 A / dm² to 0.17 A / dm² (0.03 A / dm² / μm x film thickness deviation 0.6 μm), to more accurately approach the target value.

[0124] In another application scenario, due to the increase in service life of a certain oxidation tank equipment, the temperature control delay time has increased from the historical average of 10 minutes to 15 minutes, resulting in multiple process adjustment lags. The parameter self-learning module records the temperature response feedback data of the equipment: under the same temperature adjustment instruction, the actual temperature reaches the target value 5 minutes longer than the process database preset value, and the film thickness growth rate fluctuation amplitude increases by 0.03 μm / min within 30 minutes after adjustment. The system generates a temperature response weight correction factor accordingly: the temperature response weight parameter is increased from the default value of 1.0 to 1.2 to enhance the advance of temperature adjustment. For example, when the temperature needs to be raised from 23°C to 25°C, the system issues the instruction 5 minutes in advance, and increases the adjustment amplitude from 2°C to 2.4°C (2°C x 1.2), offsetting the delay effect.

[0125] The feedback data collection of the parameter self-learning module covers the entire process chain. For the pretreatment link (such as insufficient alkali etching time leading to poor adhesion of the oxide film), the system obtains data such as the concentration, temperature, and processing time of the alkali etching solution through sensors implanted in the pretreatment equipment, and correlates it with the adhesion detection results of the film layer after oxidation (such as the grid method rating), to generate a synergistic correction factor for the pretreatment parameters and the oxidation process. For example, it is found that for every 2°C decrease in alkali etching solution temperature, the oxide film adhesion standard rate decreases by 10%, so when adjusting the oxidation process parameters, the lower limit of the current density is automatically increased by 0.1 A / dm² to compensate for the pretreatment differences.

[0126] The data security mechanism ensures the reliability of the learning process. All feedback data is stored after being de-identified and deleting sensitive information such as the device unique identifier. The update of the correction factor requires double verification: first, the system automatically performs historical data fitting verification (R² value needs to be >0.8), and second, the process engineer manually reviews the adjustment logic to avoid false learning due to abnormal data. For example, due to a sensor failure, the glossiness data jumps, and the system finds that the R² value is only 0.4 during automatic verification, so this correction factor is not adopted, and a sensor failure alarm is triggered at the same time.

[0127] The timeliness design of the parameter self-learning module is: after completing 5 batches of production, automatically start a global coefficient optimization, retrain the dynamic correction coefficient set comprehensively all feedback data. The training process uses the random forest algorithm in machine learning to extract features from more than 1000 historical data, identify key influencing factors (such as the influence weight of current density on uniformity is 45%, the influence weight of concentration on color difference is 30%), and adjust the correction priority of each parameter accordingly. For example, after global optimization, the adjustment accuracy of the current density compensation coefficient is improved from ±0.01 A / dm² / μm to ±0.005 A / dm² / μm, and the response speed of the concentration adjustment coefficient is improved by 20%.

[0128] In the cross-device learning scenario, when a new device accesses the system, the parameter self-learning module automatically retrieves the historical correction factors of the same type of device (such as the same model of oxidation tank) as the initial value to speed up the process debugging of the new device. For example, the dynamic correction coefficient set of device A (model OX-200) includes a current density compensation coefficient of -0.03 A / dm² / μm and a concentration adjustment coefficient of +1.5 g / L / ΔE. When device B (same model) accesses, the system directly copies this coefficient set to the process database of device B, and marks it as "to be verified" state. In the first 10 batches of production of device B, gradually correct the coefficients with the actual feedback data to finally form the exclusive coefficients suitable for device B.

[0129] It should be noted that, in the present text, relational terms such as first and second and the like can only be used to distinguish one entity or operation from another entity or operation, without necessarily requiring or implying any such actual relationship or order between these entities or operations. Moreover, the terms "comprises", "comprising", or any other variant thereof are intended to cover non-exclusive inclusions, so that a process, method, article, or apparatus that includes a list of elements does not only include those elements, but also includes other elements not explicitly listed, or inherent to such a process, method, article, or apparatus.

[0130] Although embodiments of the present application have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made therein without departing from the principles and spirit of the application, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An aluminum trim strip oxidation treatment real-time monitoring and process parameter self-adaptive adjustment system, characterized in that, The application relates to an aluminum decorative strip oxidation process control system based on industrial internet of things, which comprises the following modules: a process data acquisition module for acquiring process parameter data and surface state data of an aluminum decorative strip oxidation tank in real time through an industrial internet of things terminal device, analyzing and processing to generate an oxidation film growth rate characteristic value and a tank liquid dynamic parameter characteristic value, and generating a process parameter dynamic adjustment set based on an initial process parameter set stored in a process database; a parameter optimization module for processing to obtain individualized process parameters of a current batch of aluminum decorative strips based on the oxidation film growth rate characteristic value and generating a dynamic oxidation control strategy; a surface monitoring module for acquiring real-time surface quality data in an aluminum decorative strip oxidation process through a multi-spectral sensor, screening out abnormal surface data meeting a dynamic adjustment range and sending the abnormal surface data to a parameter matching module; a parameter matching module for performing multi-dimensional matching of the abnormal surface data and each parameter item in the process parameter dynamic adjustment set, generating a matching degree of real-time data and each process parameter item, and selecting a process parameter corresponding to the highest matching degree as a target adjustment parameter; a process adjustment module for receiving the target adjustment parameter and calling a preset parameter adjustment protocol in the process database to drive the oxidation tank to perform a process parameter correction operation. The matching degree of real-time data and each process parameter item specifically comprises the following steps: differential calculation of thickness deviation, uniformity change gradient and color difference fluctuation amplitude in the abnormal surface data and current dynamic range, concentration adaptability threshold and temperature optimization control value of each process parameter respectively; and generating a matching degree index of real-time data and each process parameter item based on a differential calculation result.

2. The system for real-time monitoring and self-adaptive adjustment of process parameters for aluminum trim strip oxidation treatment according to claim 1, characterized in that: The process of acquiring process parameter data of the aluminum decorative strip oxidation tank through the industrial internet of things terminal device specifically comprises the following steps: identifying a unique identification code of the oxidation tank equipment; if the equipment is a newly connected equipment, collecting basic process parameters including an initial current density, an oxidation liquid concentration reference value and a temperature control range, establishing a process characteristic node based on first-time data and performing parameter calibration to generate an oxidation film growth rate characteristic value; if the equipment is a historically connected equipment, extracting a historical process data set and a surface quality change curve of the equipment, the historical process data set including current fluctuation extreme value, concentration deviation record and temperature control delay time length, and marking the historical process data set as an initial parameter set for current process analysis.

3. The system for real-time monitoring and self-adaptive adjustment of process parameters for aluminum trim strip oxidation treatment according to claim 1, characterized in that: The process of generating the process parameter dynamic adjustment set based on the initial process parameter set stored in the process database specifically comprises the following steps: extracting an initial standard set and a dynamic correction coefficient set of each process parameter from the process database, the initial standard set including a current density safety range, a concentration fluctuation tolerance interval and a temperature regulation reference value; the dynamic correction coefficient set including a current density compensation coefficient, a concentration gradient adjustment coefficient and a temperature response weight parameter; based on a real-time tank liquid dynamic parameter characteristic value, dynamically adjusting the initial standard set of each process parameter, and recording the adjusted parameter set as the process parameter dynamic adjustment set; the process parameter dynamic adjustment set including a current dynamic range, a concentration adaptability threshold and a temperature optimization control value of each parameter item.

4. The system according to claim 3, wherein the system further comprises: a real-time monitoring device for monitoring the oxidation process of the aluminum trim strip; and a self-adaptive adjusting device for adjusting the process parameters according to the monitoring results. The process of processing to obtain individualized process parameters of a current batch of aluminum decorative strips based on the oxidation film growth rate characteristic value and generating a dynamic oxidation control strategy specifically comprises the following steps: According to the characteristic value of the oxidation film growth rate and the preset oxidation process characteristic library, mode matching is performed to determine a process parameter adjustment priority sequence; Based on the adjustment priority sequence, an adaptive control strategy including parameter trigger conditions, adjustment step rules and fault handling mechanisms is generated.

5. The system for real-time monitoring and self-adaptive adjustment of process parameters for aluminum trim strip oxidation treatment according to claim 4, characterized in that: The real-time surface quality data during the oxidation process of the aluminum trim strip is obtained by the multispectral sensor, specifically including: Monitoring the real-time spectral data stream of the oxidation film formation process, including the film thickness change value, the surface uniformity index and the color difference fluctuation value; When the real-time spectral data stream exceeds the preset qualified process range, an abnormality marker is activated and the surface data of the abnormal period is extracted as effective monitoring data.

6. The system for real-time monitoring and self-adaptive adjustment of process parameters for aluminum trim strip oxidation treatment according to claim 3, characterized in that: The process parameter corresponding to the highest matching degree is selected as the target adjustment parameter, specifically including: A matching degree sorting list of each process parameter item is established, and the parameter item corresponding to the first matching degree in the list is selected; If the first matching degree is lower than the preset adjustment trigger threshold, a backup parameter set is called and matching degree calculation is performed again.

7. The system for real-time monitoring and self-adaptive adjustment of process parameters for aluminum trim strip oxidation treatment according to claim 1, characterized in that: The matching degree index of the real-time data and each process parameter item is generated based on the difference calculation result, and the specific processing process is: The current difference, concentration gradient difference and temperature deviation are standardized by using a multi-dimensional matching algorithm to generate process parameter matching degree values ranging from 0 to 100; The closer the matching degree value is to 100, the stronger the adaptability of the real-time data to the process parameter.

8. The system for real-time monitoring and self-adaptive adjustment of process parameters for aluminum trim strip oxidation treatment according to claim 1, characterized in that: It also includes a parameter self-learning module, specifically including: Recording the oxidation film quality feedback data after each process adjustment, including the actual film thickness uniformity, surface glossiness change and color difference control effect; The feedback data is verified in reverse with the dynamic adjustment set of process parameters to generate a parameter correction factor and update the dynamic correction coefficient set of the process database.

9. The system for real-time monitoring and self-adaptive adjustment of process parameters for aluminum trim strip oxidation treatment according to claim 8, characterized in that: The parameter correction factor is generated, specifically including: Based on the deviation of the feedback data from the expected process target, a current compensation factor, a concentration adjustment factor and a temperature optimization weight are calculated; An exponential weighted average algorithm is used to dynamically smooth the historical correction factors to generate a new dynamic correction coefficient set.

Images