Full-automatic intelligent continuous plating line body detection method and device

Abstract

The application relates to the technical field of hanging plating line body detection, and discloses a full-automatic intelligent continuous hanging plating line body detection method and device, wherein the method comprises the following steps: collecting real-time multi-modal data in the running process of the hanging plating line body to obtain hanging plating line body state data and electroplating liquid multi-spectral data; performing joint feature extraction to obtain a feature vector representing the relationship between the hanging plating process and the composition of the electroplating liquid; constructing a mapping model of the hanging plating production quality and recovery parameters; executing a multi-objective optimization algorithm based on variable genetic factors and creating a hierarchical collaborative control strategy to generate linkage control parameters of the hanging plating line body production parameters and the electroplating liquid recovery equipment; and performing dynamic coordination of the hanging plating speed and the electroplating liquid treatment speed to obtain a synchronization strategy of the line body running and recovery treatment, so that the method can accurately predict the optimal recovery parameters under different working conditions and solve the problem of insufficient prediction ability of traditional fixed parameter control when facing complex and changing working conditions.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of hanging plating line detection, and particularly relates to a full-automatic intelligent continuous hanging plating line detection method and device. BACKGROUND

[0002] The traditional electroplating solution recovery method mainly relies on simple physical sedimentation and filtration technology, lacks real-time monitoring capability of the electroplating line operation state and the electroplating solution composition, and causes the recovery process and the production process to be mutually separated. This separation not only makes the recovery efficiency low, but also cannot accurately control different electroplating conditions and electroplating solution composition changes, causing electroplating quality fluctuations and energy waste. Especially in the precious metal electroplating process, due to the lack of accurate composition detection and control means, the recovery process often has problems of low efficiency and poor precision, which is difficult to meet the needs of high-quality electroplating production.

[0003] The existing hanging plating line control system mainly adopts a fixed parameter control mode, which is difficult to cope with the dynamic changes of the electroplating solution composition and the process fluctuations caused by workpiece switching. At the same time, the electroplating solution recovery equipment and the hanging plating line lack effective coordination mechanisms, and the two independently run, cannot realize the optimal allocation of resources and efficient use of energy. In addition, the traditional control algorithm lacks self-adaptive ability and cannot dynamically adjust the control strategy according to the actual production situation, resulting in system response lag and insufficient control precision. SUMMARY

[0004] The present application provides a full-automatic intelligent continuous hanging plating line detection method and device, which realizes accurate prediction of optimal recovery parameters under different conditions, and solves the problem of insufficient prediction ability of traditional fixed parameter control in the face of complex changing conditions.

[0005] In the first aspect, the present application provides a full-automatic intelligent continuous hanging plating line detection method, which comprises:

[0006] Real-time multi-modal data acquisition is performed on the running process of the hanging plating line to obtain hanging plating line state data and electroplating solution multi-spectral data;

[0007] Joint feature extraction is performed on the hanging plating line state data and the electroplating solution multi-spectral data to obtain a feature vector representing the relationship between the hanging plating process and the electroplating solution composition;

[0008] Based on the feature vector, a correlation function between the hanging plating line running state and the electroplating solution recovery efficiency is constructed to obtain a mapping model of the hanging plating production quality and the recovery parameters;

[0009] According to the mapping model, a multi-objective optimization algorithm based on variable genetic factors is executed, and a hierarchical collaborative control strategy is created, to generate linkage control parameters of the production parameters of the hanging plating line body and the electroplating solution recovery equipment;

[0010] Based on the linkage control parameters, dynamic coordination of the hanging plating speed and the electroplating solution processing speed is executed, to obtain a synchronization strategy of the line body operation and the recovery processing.

[0011] In a second aspect, the present application provides a full-automatic intelligent continuous hanging plating line body detection device, which comprises:

[0012] A data acquisition module is configured to acquire real-time multi-modal data during the operation of the hanging plating line body, to obtain hanging plating line body state data and electroplating solution multi-spectral data.

[0013] A feature extraction module is configured to extract features from the hanging plating line body state data and the electroplating solution multi-spectral data, to obtain a feature vector representing the relationship between the hanging plating process and the electroplating solution composition.

[0014] A construction module is configured to construct a correlation function between the hanging plating line body operation state and the electroplating solution recovery efficiency based on the feature vector, to obtain a mapping model of the hanging plating production quality and the recovery parameters.

[0015] A multi-objective optimization module is configured to execute a multi-objective optimization algorithm based on variable genetic factors according to the mapping model, and to create a hierarchical collaborative control strategy, to generate linkage control parameters of the production parameters of the hanging plating line body and the electroplating solution recovery equipment.

[0016] A dynamic coordination module is configured to execute dynamic coordination of the hanging plating speed and the electroplating solution processing speed based on the linkage control parameters, to obtain a synchronization strategy of the line body operation and the recovery processing.

[0017] The technical scheme provided by the present application realizes all-around monitoring of the running state of the hanging plating line body through the cooperative work of various sensing devices such as a high-resolution visual sensor, a current density sensor, an infrared thermal imager and a vibration sensor, and realizes accurate detection of the composition of the electroplating solution in combination with ultraviolet-visible spectroscopy, near-infrared spectroscopy and Raman spectroscopy, and through joint feature extraction of the state data of the hanging plating line body and the multispectral data of the electroplating solution, a feature vector of the relationship between the plating process and the composition of the electroplating solution is established, the complex relationship between the electroplating quality features, uniformity features and the composition of the electroplating solution is realized, and the limitation of traditional methods that can only be analyzed singly is broken through. The correlation function between the running state of the hanging plating line body and the recovery efficiency of the electroplating solution is constructed based on the feature vector, and through Bayesian sparse parameter polynomial regression and working condition adaptive correlation function, the optimal recovery parameters under different working conditions are accurately predicted, and the problem that the traditional fixed parameter control has insufficient prediction ability when facing complex changing working conditions is solved. By introducing a variable genetic factor multi-objective optimization algorithm, the crossover probability and mutation probability of the genetic algorithm are dynamically adjusted, the balance optimization between the electroplating quality, recovery efficiency and energy consumption is realized, the defect that the traditional optimization method is easy to fall into local optimum is avoided, and the global optimization ability of the system is improved. Through the construction of the three-layer control architecture of the hanging plating line body control layer, the electroplating solution recovery equipment control layer and the global resource scheduling layer, the cooperative control of the production parameters and the recovery parameters is realized, the problem that each subsystem independently runs and is mutually fragmented in the traditional control method is solved, and the cooperative efficiency of the overall system is improved. Based on the line speed-flow matching data, the dynamic coordination of the hanging plating speed and the electroplating solution processing speed is realized, through switching and smooth transition scheme and concentration adaptive processing scheme, the system can flexibly cope with various working conditions such as workpiece switching, line speed change and shutdown, the adaptability of the system to the changing environment is enhanced, and the overall stability of the electroplating process and the recovery process is improved.

[0018] Other features and advantages of the present application will be set forth in the following description, and in part will become apparent to those skilled in the art from the description, or can be learned by practice of the present application. The objects and other advantages of the present application will be realized and attained by the structure particularly pointed out in the description, claims and drawings.

[0019] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the following preferred embodiments are specifically described below, and the accompanying drawings are shown as follows. BRIEF DESCRIPTION OF DRAWINGS

[0020] Figure 1 An embodiment schematic diagram of the full-automatic intelligent continuous hanging plating line body detection method in the embodiments of the present application;

[0021] Figure 2 An embodiment schematic diagram of the full-automatic intelligent continuous hanging plating line body detection device in the embodiments of the present application. DETAILED DESCRIPTION

[0022] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions of the present application will be described clearly and completely below with reference to the drawings. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the present application.

[0023] The terms "comprise" and "have" and any variations thereof mentioned in the embodiments of the present application are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units is not limited to the listed steps or units, but optionally further comprises other steps or units not listed, or optionally further comprises other steps or units inherent to the process, method, product or device.

[0024] In order to facilitate the understanding of the embodiments, first, a full-automatic intelligent continuous hanging plating line body detection method disclosed by the embodiments of the present application will be described in detail. As shown in Figure 1 The method comprises the following steps:

[0025] 101. Real-time multi-modal data acquisition is performed on the running process of the hanging plating line body to obtain hanging plating line body state data and electroplating solution multi-spectral data;

[0026] It can be understood that the execution subject of the present application can be a full-automatic intelligent continuous hanging plating line body detection device, and can also be a terminal or a server, and the specific execution subject is not limited here. The embodiments of the present application take the server as the execution subject for example.

[0027] Specifically, a high-resolution vision sensor with a resolution exceeding 4K captures real-time images of the electroplated workpiece surface on the plating line, identifying key features such as microscopic defects, coating thickness variations, and adhesion uniformity. Simultaneously, image enhancement algorithms are used to denoise, sharpen, and correct illumination in the acquired images, yielding complete workpiece surface image data. Meanwhile, current density sensors continuously monitor the current distribution within the electroplating tank. These sensors are evenly distributed across key areas of the tank, capturing real-time spatial variations in current density. Time-series analysis is then used to obtain current density distribution data, ensuring timely monitoring of current uniformity and abnormal fluctuations during the electroplating process. Finally, an infrared thermal imager dynamically scans the temperature field of the plating line. This high-sensitivity, wide-band response imager accurately monitors temperature changes in the plating tank, workpiece surface, and various parts of the plating line, generating temperature field mapping data. The mechanical operation of the plating line is continuously monitored using multi-channel vibration sensors. These sensors are installed on key moving parts of the plating line, including drive motors, chains, and conveyor devices. By analyzing the frequency characteristics, amplitude variations, and harmonic components of the vibration signals, vibration characteristic data is generated, effectively identifying abnormal states or potential faults in the mechanical system. Multimodal fusion of workpiece surface image data, current density distribution data, temperature field mapping data, and vibration characteristic data is performed using time synchronization and spatial registration algorithms. Time synchronization uses high-precision clock signals to mark the acquisition timestamps of each data source, achieving microsecond-level synchronization. Spatial registration employs feature matching and geometric transformation algorithms to ensure consistent spatial correspondence between different data sources, generating plating line status data. Multi-band spectral acquisition of the electroplating solution is performed using ultraviolet-visible spectral sensors, near-infrared spectral sensors, and Raman spectral sensors. The ultraviolet-visible spectroscopy sensor covers the 200-800 nm wavelength range, enabling precise identification of changes in metal ion concentration; the near-infrared spectroscopy sensor covers the 800-2500 nm wavelength range, capable of detecting the concentration of organic additives, complexing agents, and impurities; while the Raman spectroscopy sensor analyzes changes in molecular structure and chemical bonds in the electroplating solution through a laser scattering mechanism. A multi-channel fiber optic coupler ensures efficient contact between the spectral sensors and the electroplating solution, with a sampling frequency of up to 10 times per second, ensuring high spatiotemporal resolution of the spectral data. The acquired multispectral data of the electroplating solution, after preprocessing including noise reduction, baseline correction, and spectral normalization, is stored together with the plating line status data in an industrial-grade database.

[0028] 102. Joint feature extraction is performed on the state data of the plating line and the multispectral data of the electroplating solution to obtain a feature vector characterizing the relationship between the plating process and the composition of the electroplating solution.

[0029] Specifically, the workpiece surface image data in the hanging plating line state data is subjected to electroplating defect recognition. The workpiece surface image obtained by the high-resolution visual sensor is subjected to feature extraction by a deep convolutional neural network. The network is pre-trained and subjected to transfer learning using an electroplating surface defect image library to accurately recognize plating layer surface defects, including pinhole, pimple, blister, scorch, peeling and uneven thickness, and calculate characteristic parameters such as defect area, shape, distribution density and depth to form an electroplating quality feature set. Meanwhile, the current density distribution data in the hanging plating line state data is subjected to electrofield uniformity calculation by solving a three-dimensional electrofield distribution model. The model is based on the finite element method, combined with the geometry and electrode arrangement of the electroplating tank, to construct an electrofield distribution simulation model. The current density distribution data is generated by boundary condition constraint and multiple iteration calculation. The uniformity of the electrofield distribution is quantified by the mean square deviation method and the current density deviation index, and key parameters such as the maximum current density, the minimum current density, the standard deviation and the uniformity coefficient are extracted to form an electroplating uniformity feature set for representing the consistency of the current distribution in the hanging plating process. Spectral feature extraction is performed on the electroplating solution multi-spectral data. The multi-dimensional data of ultraviolet-visible spectrum, near-infrared spectrum and Raman spectrum are combined to construct a spectral feature extraction framework based on a multi-layer convolutional neural network and a long short-term memory network. The original spectral data is preprocessed, including baseline correction, denoising, standardization and dimension compression. The local features of the spectral signal are extracted by the convolution layer, and the time sequence related features are extracted by the long short-term memory network layer to obtain the composition feature set of the electroplating solution, which contains multi-dimensional information such as metal ion concentration, organic additive content and impurity level. The quality-composition correlation model is established by time sequence correlation analysis of the electroplating quality feature set and the electroplating solution composition feature set. The dynamic time warping algorithm is used to time sequence match the quality features and composition features at different time points in the electroplating process, and the correlation coefficient matrix is calculated to identify the feature change trend of the key time sequence points, thereby extracting the quality-composition correlation features. Meanwhile, the electroplating uniformity feature set and the electroplating solution composition feature set are subjected to spatial correlation analysis. The hotspot mapping algorithm is used to divide the electroplating tank area into multiple micro-units, and the current density uniformity features and the electroplating solution composition features of different micro-units are subjected to spatial correlation calculation based on spatial statistics. The key areas of uneven current distribution or abnormal electroplating solution composition in the electroplating tank are identified by the hotspot mapping algorithm to form the uniformity-composition correlation features. The quality-composition correlation features and the uniformity-composition correlation features are fused to form a feature vector representing the relationship between the hanging plating process and the electroplating solution composition by using a multi-modal feature fusion algorithm such as self-attention mechanism or graph neural network (GNN). The feature vector has a length of 128 and contains the workpiece surface quality, current uniformity, electroplating solution composition and their time sequence and spatial correlation features in the hanging plating process.

[0030] 103. Constructing a correlation function between the running state of the hanging plating line and the recovery efficiency of the electroplating solution based on the feature vector, to obtain a mapping model of the hanging plating production quality and the recovery parameter;

[0031] Specifically, the feature vectors are subjected to nonlinear dimensionality reduction processing, and through t-distributed Stochastic Neighbor Embedding or Kernel Principal Component Analysis, high-dimensional feature data is mapped to a low-dimensional embedding space. In the dimensionality reduction process, the key information in the feature vectors is preserved, and redundant dimensions and noise interference are effectively removed, obtaining the plating-recovery feature dimensionality reduction data. A multi-input multi-output relationship network is constructed using the plating-recovery feature dimensionality reduction data. This network adopts a multi-layer perceptron structure, and outputs the plating solution recovery efficiency and related recovery parameters by inputting the multi-modal feature dimensionality reduction data of the plating line. On this basis, a Bayesian sparse parameter polynomial regression is adopted, a first-order to third-order polynomial basis function set is constructed, a Gaussian scale mixture prior is introduced, and a sparsity constraint is set for the model parameters, so as to realize the accurate mapping between the feature vectors and the recovery efficiency. The model parameters are optimized by maximizing the lower bound of the evidence, and adaptive L1 regularization is adopted to suppress the polynomial coefficients of non-key terms, so as to obtain a preliminary correlation function. For the preliminary correlation function, a working condition classification mechanism is introduced, the typical operating conditions of the plating line are identified through a K-means clustering algorithm, and the operating data are divided into different working condition categories in combination with the current density distribution, plating solution temperature, metal ion concentration and plating speed and other key parameters. For each working condition, a sub-model target training mechanism is adopted to independently model and train the feature dimensionality reduction data of each working condition category, forming a working condition adaptive correlation function, so as to ensure that the model can be optimized for different operating conditions, and the prediction accuracy and model robustness are improved. On this basis, a bidirectional mapping model of plating quality and recovery parameters is constructed. The model combines the working condition adaptive correlation function, and through a backward reasoning algorithm, the optimal recovery parameters are deduced from the target quality. The backward reasoning adopts a backward optimization process based on gradient descent. First, the target plating layer quality standard is fixed, the loss function of the deviation of the plating solution composition from the target concentration is minimized, and the recovery parameters are gradually adjusted until the recovery liquid concentration and the target quality requirement are consistent, and finally the recovery parameter prediction model is obtained. The recovery parameter prediction model is combined with the real-time plating line speed to construct a speed collaborative response model. The model adopts a dynamic time series modeling method to model the nonlinear relationship between the plating line speed change and the recovery rate, and realizes the synchronous control between the two. The speed collaborative response model can automatically adjust the recovery rate according to the change of the plating line speed, so that the composition of the plating solution is maintained within the optimal range, and the fluctuation of the plating quality caused by the speed change is prevented. The speed collaborative response model is combined with the plating tank liquid level control strategy to constrain the liquid level balance. The liquid level balance control adopts an adaptive PID controller to automatically adjust the liquid supplement and recovery rate according to the dynamic balance relationship between the plating line speed, recovery rate and plating solution replenishment rate, so as to ensure that the plating solution liquid level is always maintained within a safe range, and finally the mapping model of the plating production quality and the recovery parameters is obtained.

[0032] 104. Perform the variable genetic factor based multi-objective optimization algorithm according to the mapping model and create the hierarchical collaborative control strategy, generate the linkage control parameters of the plating line body production parameters and the electroplating liquid recovery equipment;

[0033] Specifically, the mapping model of the hanging plating production quality and the recovery parameters is converted into a multi-objective optimization problem, and the optimization objectives include maximizing the hanging plating line production quality, improving the efficiency of the electroplating solution recovery, minimizing energy consumption, and enhancing system stability. Meanwhile, the historical data and process expert rules are combined to dynamically adjust the weights of each objective. To improve the optimization efficiency, a multi-objective optimization algorithm with variable genetic factors is used for solving. This algorithm introduces adaptive mutation rate and crossover rate based on traditional genetic algorithm, and dynamically adjusts the genetic factors according to the population evolution stage, thereby improving the global search ability and convergence speed. The target weight configuration is based on the Pareto frontier analysis, which dynamically adjusts the weight distribution of the objective function and combines the weighted summation method to normalize different objectives, so as to find the optimal solution in the multi-objective space. Based on the target weight configuration, the control layer of the hanging plating line is constructed, which is based on the feature vector and optimization parameters extracted from the mapping model, and uses the model predictive control method to generate the hanging plating production control instructions, including the hanging plating line speed, hanging angle, current density, immersion time and other key parameters, to realize precise control of the production process through rolling time domain optimization. At the same time, according to the mapping model, the control layer of the electroplating solution recovery equipment is constructed, which combines the Bayesian sparse parameter polynomial regression model to calculate the control instructions of the recovery equipment, including the self-priming pump flow, centrifuge speed, heating coil power, electric push rod position and other key control variables, and dynamically adjusts through real-time monitoring data to ensure the recovery equipment in the best operating state. The cooperative execution mechanism of the hanging plating production control instructions and the recovery equipment control instructions is designed, which ensures the linkage and safety of the hanging plating and recovery process through the establishment of parameter change transfer function and interlock protection logic. The parameter change transfer function is based on a multi-input multi-output system, which constructs a dynamic correlation equation between the hanging plating speed, current density, temperature change and the recovery rate, liquid level change, and realizes real-time parameter adjustment through the state feedback mechanism. The interlock protection logic uses finite state machine for management, sets up three running states of normal, alarm and fault, triggers different levels of protection strategies in different states, and prevents the system from losing control due to parameter mutation or abnormal conditions. The cooperative mechanism ensures that the hanging plating production and the recovery equipment realize dynamic matching of parameters in each process stage, thereby optimizing the overall performance of the system. At the same time, the global resource scheduling layer of the hanging plating line and the recovery equipment is constructed, which takes the multi-objective optimization results as input, combines the real-time monitoring of the equipment state, production task priority, energy consumption limit and other constraint conditions, and uses the strategy of combining integer linear programming and reinforcement learning to solve the overall optimization problem under resource constraints to get the optimal allocation scheme. The hierarchical collaborative control strategy and the optimal allocation scheme are integrated, and the control conversion logic of different production stages is managed through the state machine.The state machine management establishes a state transition matrix for each production stage, defines the control logic of different stages such as plating starting, accelerating, stabilizing, decelerating and stopping, and triggers state transition according to real-time monitoring data, so as to realize dynamic switching and collaborative adjustment of control parameters, and finally obtain the linkage control parameters of the plating line body production parameters and the electroplating solution recovery equipment.

[0034] According to the mapping model of plating production quality and recovery parameters, different electroplating quality indicators are converted into absolute error functions. The recovery efficiency indicator is converted into a relative efficiency function, taking the extraction rate of metal ions in the electroplating solution or the concentration change rate of key components in the electroplating solution as the relative indicator of recovery efficiency, and the energy consumption indicator is converted into a unit cost function. The unit energy consumption cost is calculated by combining the energy consumption per kilowatt-hour in the recovery process with the unit recovery amount, forming a multi-objective function set. According to the mapping model, the hanging plating line control parameters (including hanging plating speed, current density, hanging angle, immersion time, etc.) and electroplating solution recovery equipment control parameters (including self-priming pump flow, centrifuge speed, heating coil power, electric push rod position, etc.) are uniformly coded into a multi-dimensional vector. This coding uses real number coding or binary coding, depending on the value range and accuracy requirements of the parameters, and the physical constraints of each control parameter (such as maximum current density, minimum flow, temperature range, etc.) are converted into variable boundary conditions, thereby defining the decision space. Process constraints are applied to the decision space, including workpiece surface quality, plating layer uniformity, recovery rate stability, etc., and a coupling constraint equation between electroplating quality and recovery efficiency is established. These coupling constraints are modeled based on the nonlinear relationship between electroplating solution composition changes, current density distribution, and recovery rate. At the same time, to prevent conflicts and unstable states between control parameters, balance constraint equations between parameters are modeled. These equations capture the dynamic balance relationship between different parameters through multivariate correlation analysis, resulting in a multi-objective optimization problem. Based on the multi-objective optimization problem, the decision variables are mapped to binary or real number chromosomes. In the genetic algorithm framework, each chromosome corresponds to a possible solution, and a fitness function is defined to reflect the goodness of the chromosome. The fitness function uses a weighted comprehensive evaluation method to weight and sum the electroplating quality error, recovery efficiency improvement rate, and energy consumption, and obtains the final fitness value according to the target weight configuration. The higher the fitness value, the better the solution. Based on this fitness function, a genetic factor adjustment mechanism for multi-objective optimization algorithms is introduced. During the multi-objective evolutionary calculation process, the crossover probability and mutation probability are dynamically adjusted. The crossover probability is positively correlated with population diversity, and the mutation probability presents a nonlinear change with the evolution generation. To prevent the algorithm from falling into local optima, the genetic factor adjustment mechanism uses an adaptive adjustment strategy to automatically adjust the crossover rate and mutation rate based on the fitness change rate, population convergence degree, and evolution stage. Adaptive genetic parameters are dynamically calculated. Using adaptive genetic parameters for multi-objective evolutionary calculation, a candidate weight configuration set is generated from the Pareto optimal solution set, which covers the optimal solution set under different target weight configurations. Through a decision preference function, candidate solutions are sorted, and combined with domain expert rules, production experience, and target priority information, the optimal weight configuration is selected, and the target weight configuration of the multi-objective optimization problem is obtained.

[0035] 105. Based on the linkage control parameter, the dynamic coordination of the plating speed and the electroplating liquid treatment speed is performed to obtain a synchronization strategy of the line operation and the recovery treatment.

[0036] Specifically, the proportional relationship between the hanging plating line speed and the electroplating solution flow rate is analyzed according to the linkage control parameters. By establishing a mapping function between the line speed and the electroplating solution flow rate, the variables such as the running speed of the hanging plating line, the current density, and the immersion time are coupled with the parameters such as the electroplating solution recovery flow rate, the self-priming pump power, and the centrifuge speed for multi-dimensional coupling analysis. The line speed-flow rate matching data are obtained, which reflects the variation law of the electroplating solution treatment speed under different hanging plating speeds. Based on the line speed-flow rate matching data, a dynamic flow rate adjustment mechanism is constructed. The model predictive control and Kalman filtering are used to quickly respond to the change of the hanging plating line speed. When the hanging plating line speed suddenly changes, the system recalculates the optimal flow rate adjustment scheme within 50 ms and dynamically adjusts the self-priming pump flow rate, the centrifuge speed, and the heating coil power, so as to keep the synchronization matching between the electroplating solution recovery equipment and the hanging plating line speed. At the same time, combined with the line body running state and the workpiece size change, a switching smooth transition scheme is designed. The S-shaped curve acceleration / deceleration algorithm is used to gradually adjust the hanging plating speed and the electroplating solution flow rate, so as to avoid the fluctuation of the plating layer thickness or the decrease of the recovery efficiency caused by the sudden change of the speed or the switching delay. The smooth transition during the switching of different batches of workpieces is effectively realized through the scheme, so as to maintain the continuity and stability of the running of the hanging plating line. At the same time, combined with the real-time monitoring results of the electroplating solution composition, an adaptive mapping relationship between the composition concentration and the treatment speed is constructed. The ultraviolet-visible spectrum, near-infrared spectrum, and Raman spectrum sensors are used to continuously monitor the key parameters such as the metal ion, the organic additive, and the impurity concentration in the electroplating solution. The self-supervised learning model is used to extract the features and predict the trends of the composition concentration changes at different times. The treatment speed of the electroplating solution is dynamically adjusted according to different concentration levels, so as to form a concentration adaptive treatment scheme. The response analysis of the recovery equipment during the planned and unplanned shutdown of the hanging plating line is carried out, and the treatment scheme during the shutdown process is obtained. For the planned shutdown, the hanging plating speed is gradually reduced before the shutdown, and the self-priming pump flow rate and the centrifuge speed are also reduced. The linear annealing algorithm is used to gradually reduce the system load, so as to ensure the smooth progress of the shutdown process. For the unplanned shutdown, an emergency response mechanism is triggered. The fast fault diagnosis model is used to identify the shutdown cause, and the priority drainage and electroplating solution cooling strategies are executed, so as to prevent the deterioration of the electroplating solution composition during the shutdown period or the abnormal state of the electroplating solution recovery equipment. In order to prevent the imbalance of the electroplating solution concentration caused by long-time shutdown, the concentration balancing strategy is automatically executed. When the shutdown is restored, the recovery flow rate and the heating coil power are preferentially adjusted, so that the electroplating solution quickly returns to the stable state. By integrating the dynamic flow rate adjustment mechanism, the switching smooth transition scheme, the concentration adaptive treatment scheme, and the shutdown process treatment scheme, the synchronization strategy of the line body running and the recovery treatment is formed.The synchronization strategy adopts a hierarchical cooperative control mechanism for management, realizes real-time optimization through a global resource scheduling layer, automatically adjusts control parameters in different production stages, to realize dynamic matching between plating speed, electroplating solution recovery speed and component concentration, and switches control logic of different process stages through a state machine control model, to ensure that the system can realize efficient cooperation of the plating and recovery processes under various working conditions.

[0037] In the embodiment of the present application, through the cooperative work of various sensing devices such as high-resolution visual sensors, current density sensors, infrared thermographs and vibration sensors, the running state of the plating line is monitored comprehensively, and the composition of the electroplating solution is accurately detected by combining ultraviolet-visible spectrum, near-infrared spectrum and Raman spectrum. Through joint feature extraction of the plating line state data and the electroplating solution multi-spectral data, the feature vector of the relationship between the plating process and the electroplating solution composition is established, the complex relationship between the electroplating quality features, uniformity features and the electroplating solution composition is deeply understood, and the limitation of traditional methods that can only be analyzed singly is broken through. Based on the feature vector, the correlation function between the running state of the plating line and the recovery efficiency of the electroplating solution is constructed, and through Bayesian sparse parameter polynomial regression and working condition adaptive correlation function, the optimal recovery parameters under different working conditions are accurately predicted, solving the problem of insufficient prediction ability of traditional fixed parameter control when facing complex changing working conditions. By introducing the variable genetic factor multi-objective optimization algorithm, the crossover probability and mutation probability of the genetic algorithm are dynamically adjusted to realize the balanced optimization of electroplating quality, recovery efficiency and energy consumption, avoiding the defect that the traditional optimization method is easy to fall into local optimum, and improving the global optimization ability of the system. Through the construction of the three-layer control architecture of the plating line control layer, the electroplating solution recovery equipment control layer and the global resource scheduling layer, the cooperative control of production parameters and recovery parameters is realized, solving the problem of independent operation and mutual fragmentation of each subsystem in traditional control methods, and improving the cooperative efficiency of the overall system. Based on the line speed-flow matching data, the dynamic coordination of the plating speed and the electroplating solution processing speed is realized, through switching the smooth transition scheme and the concentration adaptive processing scheme, the system can flexibly cope with various working conditions such as workpiece switching, line speed change and shutdown, enhancing the adaptability of the system to changing environment and improving the overall stability of the plating process and the recovery process.

[0038] In a specific embodiment, the process of step 101 can specifically include the following steps:

[0039] The high-resolution visual sensor is used to collect real-time images of the surface of the electroplated workpiece on the plating line, to obtain workpiece surface image data, and the current density sensor is used to continuously monitor the current distribution in the electroplating tank of the plating line, to obtain current density distribution data;

[0040] An infrared thermal imager is used to dynamically scan the temperature field of the plating line body to obtain temperature field mapping data, and a multi-channel vibration sensor is used to continuously monitor the mechanical operation state of the plating line body to obtain vibration characteristic data.

[0041] The workpiece surface image data, the current density distribution data, the temperature field mapping data and the vibration characteristic data are time-synchronized and spatially registered to obtain plating line body state data.

[0042] An ultraviolet-visible spectrum sensor, a near-infrared spectrum sensor and a Raman spectrum sensor are used to collect multi-band spectrum of the electroplating solution to obtain electroplating solution multi-spectrum data.

[0043] Specifically, a high-resolution visual sensor is used to collect real-time images of the surface of the electroplated workpiece on the hanging plating line. The resolution of the high-resolution visual sensor is not less than 4K, the frame rate is above 30fps, and a CMOS or CCD image sensor is used to ensure that high-definition image data of the electroplated workpiece surface can be captured during the high-speed operation of the hanging plating line. To improve image quality, the visual sensor is configured with adaptive exposure and automatic gain control algorithms to dynamically adjust exposure time and gain parameters according to the running speed of the hanging plating line and the reflection characteristics of the workpiece surface, thereby avoiding overexposure or underexposure problems. At the same time, a multi-angle lighting system provides uniform lighting conditions, and image enhancement algorithms such as histogram equalization and gamma correction are used to preprocess the collected image data, generating workpiece surface image data containing workpiece surface coating thickness, coating uniformity, and defect distribution (such as pinholes, pitting, blistering, and peeling). At the same time, current density sensors are used to continuously monitor the current distribution in the electroplating tank of the hanging plating line. These current density sensors use Hall effect or distributed current probe technology to place multiple channel current density collection points at key locations in the electroplating tank, with a sampling frequency of above 100Hz, and through multi-node signal fusion technology, global data of current density distribution is formed. To eliminate noise interference in the current signal, the collected current density data is subjected to low-pass filtering and signal noise reduction processing, and combined with the current density distribution model and the electric field simulation model, the current density is spatially interpolated and distributed reconstructed by the finite element method, generating current density distribution data. At the same time, an infrared thermal imager is used to dynamically scan the temperature field of the hanging plating line. The infrared thermal imager has a resolution of not less than 640x480, a temperature measurement accuracy of within ±0.5°C, and covers an infrared wavelength range of 8-14μm, and can capture temperature changes in the hanging plating line, workpiece surface, and electroplating tank in real time. The infrared thermal imager uses a thermal field dynamic mapping algorithm, combined with the running track of the hanging plating line and the time stamp for thermal field data calibration, and uses Kalman filtering for time series data smoothing processing to obtain temperature distribution data of the temperature field at different times and different positions, forming temperature field mapping data. To monitor the mechanical running state of the hanging plating line in real time, multiple channel vibration sensors are arranged. These sensors are installed on key moving parts of the hanging plating line (such as drive motors, chains, rollers, etc.), and the vibration sensors use MEMS accelerometers or piezoelectric sensors with a sampling frequency of above 10kHz to capture the amplitude, frequency, and harmonic characteristics of the mechanical vibration signals during the running of the hanging plating line. Through fast Fourier transform, the vibration signal is analyzed in the frequency domain, and combined with time-frequency joint analysis methods such as wavelet transform or Hilbert-Huang transform, the characteristic parameters of the vibration signal such as vibration amplitude, harmonic energy, and vibration frequency change are extracted, forming vibration characteristic data. After collecting the workpiece surface image data, current density distribution data, temperature field mapping data, and vibration characteristic data, the multi-modal data is fused through time synchronization and spatial registration algorithms.The time synchronization adopts a high-precision clock synchronization technology to ensure that the collection timestamps of each data source are accurate to the microsecond level. The spatial registration unifies the spatial coordinate systems of different data sources through feature point matching and spatial mapping transformation, and uses the iterative closest point algorithm to accurately align the data to generate the plating line body state data under a unified space-time reference. Meanwhile, the electroplating solution is collected by multi-band spectral sensors, including an ultraviolet-visible spectral sensor covering a 200-800 nm band, a near-infrared spectral sensor covering an 800-2500 nm band, and a Raman spectral sensor. The ultraviolet-visible spectral sensor can analyze the concentration changes of metal ions in the electroplating solution, the near-infrared spectral sensor can detect the concentration changes of organic additives and complexing agents in the electroplating solution, and the Raman spectral sensor can accurately detect the chemical structure changes of the electroplating solution components by identifying the vibration modes of different molecular bonds in the electroplating solution through a laser scattering mechanism. The multi-channel spectral acquisition system uses fiber coupling and signal beam splitting technology to ensure that spectral data of different bands can be collected synchronously, and the multi-spectral data of the electroplating solution are obtained through spectral standardization and baseline correction algorithms for data preprocessing.

[0044] In a specific embodiment, the process of performing step 102 can specifically include the following steps:

[0045] The workpiece surface image data in the plating line body state data are subjected to electroplating defect recognition and defect feature parameter extraction to obtain an electroplating quality feature set;

[0046] The current density distribution data in the plating line body state data are subjected to electroplating field uniformity calculation to obtain an electroplating uniformity feature set;

[0047] The electroplating solution multi-spectral data are subjected to spectral feature extraction to obtain an electroplating solution component feature set;

[0048] The time sequence correlation between the electroplating quality feature set and the electroplating solution component feature set is analyzed to obtain quality-component correlation features;

[0049] The electroplating uniformity feature set and the electroplating solution component feature set are subjected to spatial correlation analysis, and the key influence areas are identified through a hot spot mapping algorithm to obtain uniformity-component correlation features;

[0050] The quality-component correlation features and the uniformity-component correlation features are fused to form a feature vector representing the relationship between the plating process and the electroplating solution components.

[0051] Specifically, the workpiece surface image data in the hanging plating line state data is used for electroplating defect recognition and defect feature parameter extraction. The surface image data collected by a high-resolution visual sensor is combined with a deep convolutional neural network for feature extraction. The deep convolutional neural network uses ResNet50 or VGG16 as the backbone network, extracts texture features, color distribution, morphological structure, and other multi-dimensional information on the plating surface through multiple convolution kernels, and uses attention mechanisms to enhance the focusing ability on defect areas. Different defect types are classified and recognized, including pinholes, pitting, blistering, scorching, peeling, and uneven plating thickness. Morphological analysis and image segmentation algorithms such as U-Net are used to extract the boundaries and feature parameters of the defect area, including defect area, perimeter, shape coefficient, density distribution, maximum / minimum size, and depth information, to form an electroplating quality feature set. The current density distribution data in the hanging plating line state data is used for electroplating field uniformity calculation. A three-dimensional electroplating field simulation model is constructed using the finite element method. The model is based on key parameters such as electroplating tank geometry, anode-cathode distance, and spatial distribution of current density measurement points for boundary constraints, and is dynamically calibrated using multi-point current data collected by current density sensors. The electric field distribution is calculated by solving the Poisson equation, and the uniformity of the electric field distribution is quantitatively analyzed using the mean square deviation method, uniformity coefficient, and standard deviation indicators. Key features such as maximum, minimum, mean, uniformity index, and abnormal current interval are extracted to form an electroplating uniformity feature set, reflecting the uniformity of the electric field distribution and the distribution of abnormal areas during electroplating. While extracting electroplating quality and uniformity features, multispectral data of the electroplating solution are used for spectral feature extraction. Ultraviolet-visible, near-infrared, and Raman spectroscopy are used to collect electroplating solution composition data, and a spectral feature extraction framework combining long short-term memory networks and convolutional neural networks is used for deep feature extraction. The original spectral data is preprocessed through baseline correction, denoising, and dimension compression, and then local spectral features are extracted using a convolutional neural network, and temporal features are extracted using a long short-term memory network to generate an electroplating solution composition feature set containing metal ion concentration, organic additive content, impurity level, and complexing agent distribution, reflecting the trend of electroplating solution composition changes. The temporal correlation between the electroplating quality feature set and the electroplating solution composition feature set is analyzed, and the electroplating solution composition changes and workpiece surface quality features at different time periods are matched using dynamic time warping algorithm. The time lag effect of electroplating solution composition changes on plating layer quality is quantified using a correlation matrix, and the direct impact of electroplating solution composition changes on specific defect types is evaluated using Granger causality analysis to obtain quality-composition correlation features, revealing the dynamic impact mechanism of electroplating solution composition fluctuations on plating layer quality changes. Meanwhile, spatial correlation analysis is performed on the electroplating uniformity feature set and the electroplating solution composition feature set, and hot spot mapping algorithm is used to identify key impact areas in the electroplating tank.The algorithm generates a three-dimensional space hotspot distribution map by spatial interpolation of the uniformity of the electric field and the concentration distribution of the electroplating solution based on the Kriging interpolation method, and combines spatial weighted clustering to evaluate the hotspot area in stages, thereby identifying the key impact area of current density unevenness and electroplating solution concentration abnormality on the uniformity of the plated layer, and extracting key feature parameters of the hotspot area, such as current deviation, local concentration abnormal area, hotspot area center coordinates, etc., to form uniformity-ingredient correlation features. These features help to accurately locate the high-risk area in the electroplating tank that causes uneven plating or defects. By fusing the quality-ingredient correlation features and the uniformity-ingredient correlation features, a multi-modal feature fusion algorithm (such as a self-attention mechanism) is used to weight the features of different modalities, and a feature selection algorithm (such as principal component analysis) is used to extract a key feature subset to form a feature vector representing the relationship between the plating process and the electroplating solution composition. The feature vector has a length of 128 dimensions, covering multi-dimensional information such as workpiece surface quality, current uniformity, electroplating solution composition changes, and key impact area distribution.

[0052] In a specific embodiment, the process of performing step 103 can specifically include the following steps:

[0053] Nonlinear dimensionality reduction is performed on the feature vector to obtain plating-recovery feature dimensionality reduction data;

[0054] A multi-input multi-output relationship network is constructed using the plating-recovery feature dimensionality reduction data, and a mapping relationship between the feature vector and the recovery efficiency is established through Bayesian sparse parameter polynomial regression to obtain a preliminary correlation function;

[0055] A plating line body operating condition classification mechanism is introduced for the preliminary correlation function, and a clustering algorithm is used to identify typical operating conditions and train a target sub-model for each type of operating condition to obtain a condition-adaptive correlation function;

[0056] Based on the condition-adaptive correlation function, a bidirectional mapping between electroplating quality and recovery parameters is constructed, and a reverse reasoning algorithm is used to deduce the optimal recovery parameters from the target quality to obtain a recovery parameter prediction model;

[0057] A speed coordination response model of line speed and recovery rate is established by combining the recovery parameter prediction model with the real-time plating line speed;

[0058] The speed coordination response model is combined with the electroplating tank liquid level control strategy to constrain the liquid level balance, and a mapping model of plating production quality and recovery parameters is obtained.

[0059] Specifically, the feature vectors formed by the plating line state data, electroplating solution composition characteristics, and electroplating uniformity characteristics are subjected to nonlinear dimension reduction. Due to the high dimensionality (128 or more) of the feature vectors, direct mapping can increase the computational complexity and slow down the model convergence. Therefore, t-distributed Stochastic Neighbor Embedding (t-SNE) or Kernel Principal Component Analysis (KPCA) is used for dimension reduction. Both of these dimension reduction methods can effectively preserve the nonlinear structure information of high-dimensional data. t-SNE maps the feature vectors to a 20-dimensional embedding space by minimizing the Kullback-Leibler (KL) divergence between the original data space and the low-dimensional space. KPCA maps high-dimensional features to low-dimensional features through a kernel function while preserving the nonlinear relationships between features. The dimension-reduced data retains the core information of the plating line running state, electroplating solution composition changes, and electroplating uniformity characteristics, generating plating-recovery feature dimension-reduced data. The plating-recovery feature dimension-reduced data is used to construct a multi-input multi-output relationship network to establish the mapping relationship between the plating line running state, plating layer quality, and recovery efficiency. The multi-input multi-output relationship network uses a multilayer perceptron structure, with the input layer accepting the dimension-reduced feature data and the output layer simultaneously generating multi-dimensional results such as recovery efficiency, plating layer uniformity, and target concentration changes. On this basis, a Bayesian sparse parameter polynomial regression model is established for the multi-input multi-output relationship network output. The Bayesian sparse parameter regression model achieves sparsity constraints on polynomial coefficients by introducing a Gaussian scale mixture prior. In the regression model, key variables are given higher weights, and the influence of irrelevant variables is reduced, resulting in a preliminary correlation function between the feature vector and the recovery efficiency. The Bayesian sparse parameter polynomial regression model optimizes parameters using the maximum evidence lower bound and iteratively updates regression coefficients through variational inference, forming a preliminary multi-dimensional correlation function. A plating line running condition classification mechanism is introduced for the preliminary correlation function. The K-means clustering algorithm is used to identify different typical working conditions, and the feature data is classified according to the working condition type. The K-means algorithm clusters the dimension-reduced data into different categories such as low current density operation, high current density operation, stable plating speed, and high temperature working conditions. A target sub-model is trained for each working condition category. Transfer learning is used to fine-tune the preliminary correlation function parameters in different working condition categories, generating a working condition adaptive correlation function. These sub-models are self-adaptively optimized according to different running states, significantly improving the model's generalization ability and accuracy. Based on the working condition adaptive correlation function, a bidirectional mapping model of electroplating quality and recovery parameters is constructed. This mapping model uses a bidirectional long short-term memory network to model the temporal correlation between the electroplating solution composition, plating speed, and plating layer quality. Through a backtracking algorithm, the optimal recovery parameters are derived from the target quality.The reverse reasoning algorithm utilizes the gradient backpropagation mechanism to perform reverse deduction on the target plating layer thickness, uniformity and other quality indicators. By minimizing the error between the target quality and the model predicted quality, the recovery parameters (such as recovery flow rate, centrifuge speed, heating coil power, etc.) are iteratively adjusted until the recovery liquid concentration and plating layer quality reach optimal matching, forming a recovery parameter prediction model. Combining the recovery parameter prediction model with the real-time hanging plating line speed, a speed coordination response model of line speed and recovery rate is established. This model is based on a nonlinear autoregressive model and dynamic time warping to model the dynamic relationship between hanging plating line speed changes and recovery rate changes. The nonlinear autoregressive model captures the lag response of hanging plating speed changes to recovery rate, and combines dynamic time warping to match the recovery rate trend under different speed change patterns, forming a speed coordination response model that can adaptively adjust. According to the fluctuations in the hanging plating speed, the control parameters of the recovery equipment are automatically adjusted to ensure that the plating layer quality and recovery efficiency remain in an optimal matching state. In order to improve the stability of the hanging plating process and the electroplating liquid recovery process, the speed coordination response model is integrated with the electroplating tank liquid level control strategy, and a liquid level balance constraint mechanism is introduced to maintain the dynamic balance of the electroplating tank liquid level. The liquid level balance control strategy uses an adaptive PID controller to automatically adjust the liquid supplement rate and recovery rate based on the dynamic balance relationship between the recovery rate change, the hanging plating speed change and the electroplating liquid supplement rate. When the liquid level fluctuates, the Kalman filter is used to predict the liquid level change, thereby ensuring that the electroplating tank liquid level is always at an optimal level. Through the introduction of the liquid level balance constraint, the abnormal situation caused by the imbalance between the recovery rate and the liquid supplement rate is effectively avoided, and multi-objective collaborative optimization of hanging plating speed, recovery rate and liquid level control is achieved. By integrating the speed coordination response model and the liquid level balance constraint mechanism, a mapping model of hanging plating production quality and recovery parameters is constructed.

[0060] In a specific embodiment, the process of performing step 104 can specifically include the following steps:

[0061] Convert the mapping model into a multi-objective optimization problem, and set the target weight configuration of the multi-objective optimization problem based on the multi-objective optimization algorithm with variable genetic factors;

[0062] Construct a hanging plating line body control layer based on the target weight configuration, and generate a hanging plating production control instruction according to the hanging plating line body control layer;

[0063] Construct an electroplating liquid recovery equipment control layer according to the mapping model, and calculate a recovery equipment control instruction through the electroplating liquid recovery equipment control layer;

[0064] Design a collaborative execution mechanism for the hanging plating production control instruction and the recovery equipment control instruction, and establish a parameter change transfer function and interlock protection logic according to the collaborative execution mechanism to obtain a hierarchical collaborative control strategy;

[0065] A global resource scheduling layer of the hanging plating line body and the recovery equipment is constructed, and a whole optimization problem under resource constraints is solved according to the global resource scheduling layer to obtain an optimal allocation scheme.

[0066] An integrated layered collaborative control strategy and the optimal allocation scheme are integrated, and control conversion logic of different production stages is managed through a state machine to obtain linkage control parameters of the hanging plating line body production parameters and the electroplating solution recovery equipment.

[0067] Specifically, the mapping model of plating production quality and recovery parameters is transformed into a multi-objective optimization problem for core indicators such as plating production quality, recovery efficiency, and energy consumption. In the mapping model, different objectives have competitive characteristics. A multi-objective optimization model is used to mathematically model different objective functions, and a variable genetic factor mechanism is introduced to optimize the target weight configuration. The variable genetic factor mechanism adjusts the crossover rate, mutation rate, and population size adaptively, dynamically adjusts the genetic algorithm parameters according to the changes in different target weights, and improves the global search ability of the optimization algorithm under different target weights. The target weight configuration is optimized based on the Pareto optimal front analysis. By balancing the mutual influence between plating production quality, plating solution recovery efficiency, and energy consumption, a set of candidate target weight configuration sets is generated. Combined with the decision preference function, the candidate solutions are screened to determine the optimal target weight configuration. Based on the target weight configuration, the plating line control layer is constructed. The control layer is designed based on model predictive control, with the target weight configuration as the input, and the key control parameters such as plating speed, current density, immersion time, and hanging angle as the control variables. The dynamic state transfer equation is used to predict and optimize the target variables such as coating thickness and uniformity. The model predictive controller updates the objective function cycle by cycle through rolling horizon optimization and adjusts the control instructions based on real-time monitoring data to generate plating production control instructions. The control instructions dynamically adapt to the changes in the running state of the plating line, achieving optimal control of coating quality and energy consumption. At the same time, the plating solution recovery equipment control layer is constructed based on the mapping model. The recovery equipment control layer uses nonlinear model predictive control for control modeling. The model takes the key control parameters of the recovery equipment such as self-priming pump flow, centrifuge speed, heating coil power, and electric push rod position as input variables. The Bayesian sparse parameter polynomial regression model is used to calculate the nonlinear mapping relationship between different recovery rates and changes in plating solution composition. Through Bayesian optimization, the recovery parameters are iteratively updated to obtain the recovery equipment control instructions. The nonlinear model predictive controller predicts the system state at multiple future times and optimizes the control variables to keep the recovery equipment in an efficient operating state, ensuring the best match between plating solution recovery efficiency and coating quality. To achieve coordinated operation of plating production and recovery equipment control, a collaborative execution mechanism for plating production control instructions and recovery equipment control instructions is designed. The collaborative mechanism transmits the changes in variables such as plating speed and current density to the recovery equipment through the construction of parameter change transmission function, and introduces interlock protection logic to prevent system instability caused by sudden changes in key control parameters. The parameter change transmission function uses multi-input multi-output dynamic modeling to build the linkage equation between the plating line and the recovery equipment, dynamically matches and adjusts the control instructions at different stages, and the interlock protection logic manages the state transition of different control stages through a finite state machine. An abnormality detection mechanism is set up to trigger the protection logic immediately if the system deviates from the safety threshold, achieving safe and collaborative control of the plating and recovery processes, and forming a hierarchical collaborative control strategy.In order to improve the efficiency of system resource utilization, a global resource scheduling layer of the plating line and the recovery equipment is constructed. The global resource scheduling layer dynamically allocates resources of the plating production and the recovery equipment through the combination of integer linear programming and reinforcement learning, and adjusts according to real-time production data. The integer linear programming model takes minimizing resource consumption and maximizing production capacity as the objective function, and the constraint conditions include device capacity limit, plating speed limit and recovery liquid level threshold, etc. The reinforcement learning continuously optimizes the scheduling strategy through adaptive learning in different resource allocation schemes by using the Q-learning algorithm. The overall optimization problem under resource constraints is solved by combining the resource scheduling result and the working condition adaptive strategy, and the optimal resource allocation scheme is obtained. After the construction of the hierarchical collaborative control strategy and the optimal allocation scheme, the hierarchical collaborative control strategy and the optimal allocation scheme are integrated, and the parameter switching and control mode conversion in different operating stages are realized through the state machine management of the control switching logic in different production stages. The state machine automatically switches different control states such as acceleration stage, stable operation stage, recovery stage and deceleration stage according to key variables such as plating line running stage, plating solution composition change and target quality deviation, and loads the corresponding control strategy in different states to realize the linkage adjustment of plating production parameters and recovery equipment control parameters. Through the dynamic switching logic of the state machine, the optimal operation of the plating line and the recovery equipment in different stages is ensured, and the problem of system instability caused by parameter mutation is avoided. A linkage control parameter system of plating line production parameters and plating solution recovery equipment is formed, which has core functions such as multi-objective optimization, adaptive control, resource scheduling and state management, and can realize efficient collaboration of plating production line and recovery equipment.

[0068] In a specific embodiment, the execution step of transforming the mapping model into a multi-objective optimization problem and setting the target weight configuration of the multi-objective optimization problem based on the variable genetic factor algorithm can specifically include the following steps:

[0069] According to the mapping model, the electroplating quality index is transformed into an absolute error function, the recovery efficiency index is transformed into a relative efficiency function, and the energy consumption index is transformed into a unit cost function to obtain a set of multi-objective functions;

[0070] According to the mapping model, the plating line control parameters and the plating solution recovery equipment control parameters are uniformly coded into a multi-dimensional vector, and the physical constraints of each parameter are transformed into variable boundary conditions to obtain a decision space;

[0071] Process constraint conditions are applied to the decision space, and coupling constraint equations between electroplating quality and recovery efficiency and balance constraint equations between control parameters are established to obtain a multi-objective optimization problem;

[0072] The decision variable is mapped to a binary or real number chromosome based on a multi-objective optimization problem, and a fitness function is defined to reflect the advantages and disadvantages of the chromosome;

[0073] A genetic factor adjustment mechanism of a multi-objective optimization algorithm is introduced based on the multi-objective optimization problem and the fitness function, and adaptive genetic parameters are obtained by dynamically calculating the crossover probability and the mutation probability based on the genetic factor adjustment mechanism;

[0074] The adaptive genetic parameters are used to perform multi-objective evolutionary calculation to obtain a candidate weight configuration set, and an optimal solution is selected based on the candidate weight configuration set by applying a decision preference function to obtain a target weight configuration of the multi-objective optimization problem.

[0075] Specifically, different objectives are mathematically modeled according to the mapping model of plating production quality and electroplating solution recovery parameters. For electroplating quality indicators, an absolute error function is used for quantification, and the quality deviation of the electroplating process is measured by comparing the gap between the target plating thickness, uniformity or adhesion rate and the actual measured value. The recovery efficiency index is modeled by a relative efficiency function, and the relative performance of the recovery equipment under different working conditions is measured by calculating the ratio of the actual recovery efficiency to the theoretical maximum recovery efficiency. At the same time, the energy consumption index is converted into a unit cost function, which compares the total energy consumption during system operation with the plating production. Through this step, a set of multi-objective functions is formed. The plating line control parameters and the electroplating solution recovery equipment control parameters are coded into a multi-dimensional vector. Continuous control variables (such as plating speed, current density, recovery flow, soaking time, etc.) are real-coded, while discrete control variables (such as equipment start-stop state, recovery mode, etc.) are binary-coded, forming a high-dimensional control vector. At the same time, in order to ensure the feasibility of the optimization solution in actual production, the physical constraints of each control parameter are converted into variable boundary conditions, such as the current density range of the electroplating tank, the flow limit of the recovery equipment, the upper limit of the heating coil power, etc. These boundary conditions constrain the optimization solution, thereby defining a multi-dimensional decision space. Process constraints are imposed on the decision space, and coupling constraint equations between electroplating quality and recovery efficiency and balance constraint equations between control parameters are established. The coupling constraint equation is used to capture the nonlinear influence of electroplating solution composition changes on plating quality and recovery efficiency, thereby ensuring that different control parameters can coordinate with each other when solving the optimization. The balance constraint equation achieves dynamic balance between the electroplating process and the recovery process by modeling the multi-variable correlation of control parameters, thereby avoiding system instability caused by excessive changes in control parameters. By imposing these constraints, the multi-objective optimization problem is converted into a multi-objective optimization problem with complex physical constraints and process limitations. Based on the multi-objective optimization problem, the decision variables are mapped to binary or real chromosomes, each chromosome representing a possible solution, and the fitness function reflects the quality of the chromosome. The fitness function combines different target weights to weight the sum of plating quality deviation, recovery efficiency and unit cost optimization targets, generating a comprehensive fitness value to guide the optimization process of the genetic algorithm. At the same time, in order to prevent the genetic algorithm from falling into local optimal solution or losing population diversity, a genetic factor adjustment mechanism of multi-objective optimization algorithm is introduced. This mechanism monitors the evolution state of the genetic algorithm in real time, dynamically adjusts the crossover probability and mutation probability according to the population convergence degree, solution distribution density and target function gradient change, and generates adaptive genetic parameters. The crossover probability controls the global search ability of the solution, and the mutation probability determines the exploration ability of the solution in the local area. Through the adaptive adjustment mechanism, the two are optimized in linkage, so that the genetic algorithm can automatically adjust the strategy at different stages to improve the solution accuracy and global convergence speed.Under the guidance of adaptive genetic parameters, the system performs multi-objective evolutionary computation, and uses the fast non-dominated sorting genetic algorithm to perform multi-objective optimization of the population. The fast non-dominated sorting genetic algorithm divides the population solutions into different levels of Pareto front through the non-dominated sorting mechanism, and combines the crowding degree sorting and elite reservation strategy to screen the optimal solution set, forming a set of candidate weight configuration set. The candidate solution set is screened through the Pareto front analysis, and the solution priority is sorted through the decision preference function, and the optimal weight configuration is selected according to the change trend of the objective function and the priority of the production process, and finally the target weight configuration of the multi-objective optimization problem is obtained.

[0076] The variable genetic factor mechanism is introduced into the basic structure of the genetic algorithm. The genetic factor adjustment function is constructed by analyzing the population diversity index and the evolution stagnation degree, including: calculating the Hamming distance or Euclidean distance for each chromosome in the population, and obtaining the average distance between population individuals through normalization processing. The population diversity index is determined by comparing the average distance with the preset threshold. The time series of the optimal fitness value of continuous generations is constructed, and the evolution stagnation degree is evaluated by calculating the mean and standard deviation of the adjacent generation fitness improvement rate. Based on the population diversity index and the evolution stagnation degree, a binary variable genetic factor adjustment function is constructed, which contains an adjustment coefficient determined according to the characteristics of the plating line body. Boundary constraint conditions are set for the binary variable genetic factor adjustment function to prevent the genetic algorithm from falling into extreme parameters, and the values of the boundary values are determined according to the special requirements of the plating line body control. The segmented adaptive crossover probability function and mutation probability function are constructed by using the variable genetic factor adjustment function with constraints, and the crossover and mutation operations are differentiated controlled according to the different characteristics of the population evolution stage. The segmented adaptive crossover probability function and mutation probability function are applied to the evolution operation of the genetic algorithm, and the genetic factor value is recalculated according to the characteristics of the new population at the end of each generation evolution, realizing the dynamic adaptive adjustment of the genetic operation.

[0077] The multi-objective evolutionary calculation is performed by using adaptive genetic parameters, and the elite individuals are selected by using the Pareto non-dominated sorting and the crowding distance calculation, including: the multi-objective function values of a plurality of chromosomes in the initial population are calculated respectively, and the population is divided into different non-dominated levels based on the Pareto dominance relationship; the crowding distance of the chromosomes in each non-dominated level is calculated, and the distribution sparseness of the individual in the solution space is obtained by summing the normalized distances of adjacent solutions in each objective function dimension; the selection operator is designed by combining the non-dominated level and the crowding distance, and the parent individuals are selected by using the binary tournament selection strategy of non-dominated level priority and crowding distance priority; the crossover operation controlled by the adaptive genetic parameters is applied to the selected parent individuals, the child chromosomes are generated by simulating binary crossover or uniform crossover, and the crossover probability is dynamically adjusted by the variable genetic factor; the mutation operation is performed on the child chromosomes after the crossover, random disturbance is introduced by polynomial mutation or Gaussian mutation, the mutation probability is inversely proportional to the variable genetic factor, and the population diversity is maintained; the parent population and the child population generated by the crossover and mutation are combined to form an intermediate population with doubled size, and the non-dominated sorting and the crowding calculation are performed again to select the optimal individuals to form a new generation population; the above evolution process is repeated until the termination condition is reached, and the individuals with the highest non-dominated level in each generation evolution are saved to an external archive set as an approximation of the Pareto optimal solution set.

[0078] In a specific embodiment, the process of performing step 105 can specifically include the following steps:

[0079] Performing proportional relationship analysis of the hanging plating line speed and the electroplating liquid flow rate according to the linkage control parameters to obtain line speed-flow rate matching data;

[0080] Using the line speed-flow rate matching data to obtain a dynamic flow rate adjustment mechanism for rapid response to changes in the hanging plating line speed, and performing transition period control for the workpiece switching process of the hanging plating line to obtain a smooth transition scheme for switching;

[0081] Establishing an adaptive mapping relationship between the component concentration and the processing speed based on the real-time monitoring results of the electroplating liquid components to obtain a concentration adaptive processing scheme;

[0082] Performing recovery equipment response analysis for planned and unplanned shutdowns of the hanging plating line to obtain a shutdown process processing scheme;

[0083] Integrating the dynamic flow rate adjustment mechanism, the smooth transition scheme for switching, the concentration adaptive processing scheme, and the shutdown process processing scheme to obtain a synchronization strategy for line operation and recovery processing.

[0084] Specifically, the proportional relationship between the hanging plating line speed and the electroplating solution flow rate is analyzed based on the linkage control parameters, and a line speed-flow rate matching model is constructed. This model establishes a matching relationship by analyzing the nonlinear relationship between the hanging plating line speed, current density, immersion time, plating layer thickness variation, and electroplating solution processing speed. The matching model uses multivariate regression analysis or support vector regression for modeling, trains through historical data, and combines the dynamic time warping algorithm to perform time series matching of electroplating solution flow rate adjustments under different speed changes, generating line speed-flow rate matching data. The matching data covers the best adjustment range of electroplating solution flow rate under different hanging plating speeds and can automatically adjust the matching parameters according to the plating layer thickness target of different batches of workpieces, ensuring that the hanging plating speed and the electroplating solution processing speed always maintain the best proportional relationship. Based on the line speed-flow rate matching data, a dynamic flow rate adjustment mechanism is constructed. This mechanism uses model predictive control and adaptive proportional-integral-derivative control to quickly respond to changes in the hanging plating line speed. When the hanging plating line speed changes, the model predictive controller calculates the optimal flow rate adjustment command within 50 ms, dynamically adjusting the flow rate of the self-priming pump, the speed of the centrifuge, and the power of the heating coil, thereby maintaining the matching relationship between the electroplating solution flow rate and the hanging plating line speed. At the same time, to deal with the transition period control during workpiece switching of the hanging plating line body, a switching smooth transition scheme is designed. This scheme uses an S-shaped speed curve to achieve smooth switching of the acceleration, constant speed, and deceleration processes, combines multi-order Bezier curve interpolation for dynamic smooth adjustment of the hanging plating line speed, and uses Kalman filtering for dynamic correction of abnormal speed changes, thereby achieving stable transition of plating layer quality and electroplating solution flow rate during workpiece switching. The switching smooth transition scheme can effectively avoid problems such as uneven plating layer thickness or fluctuation in recovery efficiency caused by sudden speed changes, thereby maintaining the stability of the hanging plating process. At the same time, an adaptive mapping relationship between the composition concentration and the processing speed is constructed based on real-time monitoring results of the electroplating solution composition. The system uses ultraviolet-visible spectrum, near-infrared spectrum, and Raman spectrum sensors for multi-band spectral analysis of the electroplating solution, combines long short-term memory networks and convolutional neural networks to extract spectral features, and uses a self-supervised learning mechanism for feature enhancement and concentration change prediction to establish an adaptive concentration processing scheme. This scheme analyzes the relationship between the best processing speed and the plating layer quality under different electroplating solution concentrations to dynamically adjust the operating parameters of the recovery equipment. When significant changes in metal ion concentration, organic additive content, or impurity level in the electroplating solution are detected, the system updates the recovery flow rate, centrifuge speed, and replenishment speed within 100 ms, thereby ensuring that the electroplating solution composition always remains within the optimal concentration range and preventing composition imbalance from adversely affecting the hanging plating quality. The response of the recovery equipment to planned and unplanned shutdowns of the hanging plating line body is analyzed, and a shutdown process processing scheme is constructed. This scheme uses a finite state machine to manage different stages of the shutdown process, including shutdown preparation, deceleration, stopping, recovery processing, and restarting.When a planned shutdown is scheduled, the system executes a linear annealing algorithm to gradually reduce the plating speed while simultaneously adjusting the plating solution recovery speed and replenishment speed, ensuring that the plating layer thickness and component concentration remain stable during the shutdown process. In the case of an unplanned shutdown, the system triggers an emergency response mechanism to identify the shutdown cause through a rapid fault detection and diagnosis model, and performs a backtracking adjustment of key control parameters through an adaptive parameter recovery algorithm, ensuring that the recovery equipment can be restored to an optimal operating state in the shortest time. At the same time, combined with the concentration balance strategy, the component concentration is rebalanced after shutdown, and when the system is restarted, the heating coil power, recovery flow rate and replenishment rate are adjusted to quickly restore the composition of the plating solution to a stable state, thereby ensuring the plating quality and recovery efficiency after the plating line is restarted. By integrating the dynamic flow adjustment mechanism, the switching smooth transition scheme, the concentration adaptive processing scheme and the shutdown process processing scheme, a synchronous strategy for line operation and recovery processing is formed.

[0085] The above describes the full-automatic intelligent continuous plating line detection method in the embodiment of the application, and the following describes the full-automatic intelligent continuous plating line detection device in the embodiment of the application, please refer to Figure 2 An embodiment of the full-automatic intelligent continuous plating line detection device in the embodiment of the application includes:

[0086] The data acquisition module 201 is configured to acquire real-time multi-modal data of the plating line during operation, and obtain plating line state data and plating solution multi-spectral data.

[0087] The feature extraction module 202 is configured to extract features from the plating line state data and the plating solution multi-spectral data, and obtain a feature vector representing the relationship between the plating process and the plating solution composition.

[0088] The construction module 203 is configured to construct a correlation function between the plating line operating state and the plating solution recovery efficiency based on the feature vector, and obtain a mapping model of the plating production quality and the recovery parameters.

[0089] The multi-objective optimization module 204 is configured to execute a multi-objective optimization algorithm based on variable genetic factors and create a hierarchical collaborative control strategy based on the mapping model, and generate linkage control parameters of the plating line production parameters and the plating solution recovery equipment.

[0090] The dynamic coordination module 205 is configured to perform dynamic coordination of the plating speed and the plating solution processing speed based on the linkage control parameters, and obtain a synchronous strategy for line operation and recovery processing.

[0091] Through the cooperation of each component, through the cooperation of various sensing devices such as high-resolution visual sensors, current density sensors, infrared thermographs and vibration sensors, the running state of the hanging plating line is monitored in all directions, and the composition of the electroplating solution is accurately detected by combining ultraviolet-visible spectrum, near-infrared spectrum and Raman spectrum. Through joint feature extraction of the hanging plating line state data and the electroplating solution multi-spectral data, the characteristic vector of the relationship between the hanging plating process and the electroplating solution composition is established, the complex relationship between the electroplating quality characteristics, uniformity characteristics and the electroplating solution composition is deeply understood, and the limitation of traditional methods that can only be analyzed singlely is broken through. Based on the characteristic vector, the correlation function between the running state of the hanging plating line and the recovery efficiency of the electroplating solution is constructed, and through the Bayesian sparse parameter polynomial regression and the working condition adaptive correlation function, the optimal recovery parameter under different working conditions is accurately predicted, and the problem of insufficient prediction ability of traditional fixed parameter control when facing complex changing working conditions is solved. By introducing the variable genetic factor multi-objective optimization algorithm, the crossover probability and mutation probability of the genetic algorithm are dynamically adjusted, the balance optimization between the electroplating quality, the recovery efficiency and the energy consumption is realized, the defect that the traditional optimization method is easy to fall into local optimum is avoided, and the global optimization ability of the system is improved. Through the construction of the three-layer control architecture of the hanging plating line control layer, the electroplating solution recovery equipment control layer and the global resource scheduling layer, the cooperative control of the production parameters and the recovery parameters is realized, the problem of independent operation and mutual fragmentation of each subsystem in the traditional control method is solved, and the cooperative efficiency of the overall system is improved. Based on the line speed-flow matching data, the dynamic coordination of the hanging plating speed and the electroplating solution processing speed is realized, through the switching smooth transition scheme and the concentration adaptive processing scheme, the system can flexibly cope with various working conditions such as workpiece switching, line speed change and shutdown, the adaptability of the system to changing environment is enhanced, and the overall stability of the electroplating process and the recovery process is improved.

[0092] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the system, the system and the unit described above can refer to the corresponding process in the foregoing method embodiments, which will not be repeated here.

[0093] The integrated unit, if implemented in the form of a software function unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application or the entire or part of the technical solutions that essentially contribute to the prior art can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in the various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0094] The above-described embodiments are merely used to illustrate the technical solutions of the present application, rather than limit the same; even though the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacements for some of the technical features; and these modifications or replacements do not cause the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A full-automatic intelligent continuous hanging plating line body detection method, characterized in that, The method comprises: real-time multi-modal data acquisition is performed on the running process of the hanging plating line body to obtain hanging plating line body state data and electroplating solution multi-spectral data; joint feature extraction is performed on the hanging plating line body state data and the electroplating solution multi-spectral data to obtain a feature vector representing the relationship between the hanging plating process and the electroplating solution composition; an association function between the hanging plating line body running state and the electroplating solution recovery efficiency is constructed based on the feature vector to obtain a mapping model of the hanging plating production quality and the recovery parameters; a multi-objective optimization algorithm based on variable genetic factors is executed according to the mapping model, and a hierarchical collaborative control strategy is created to generate linkage control parameters of the hanging plating line body production parameters and the electroplating solution recovery equipment; dynamic coordination of the hanging plating speed and the electroplating solution processing speed is performed based on the linkage control parameters to obtain a synchronization strategy of the line body running and the recovery processing.

2. The full-automatic intelligent continuous hanging plating line body detection method according to claim 1, characterized in that, The real-time multi-modal data acquisition performed on the running process of the hanging plating line body to obtain hanging plating line body state data and electroplating solution multi-spectral data comprises: real-time imaging acquisition is performed on the surface of the electroplated workpiece on the hanging plating line body by using a high-resolution visual sensor to obtain workpiece surface image data, and continuous monitoring of the current distribution in the electroplating tank of the hanging plating line body is performed by using a current density sensor to obtain current density distribution data; dynamic scanning of the temperature field of the hanging plating line body is performed by using an infrared thermal imager to obtain temperature field mapping data, and continuous monitoring of the mechanical running state of the hanging plating line body is performed by using a multi-channel vibration sensor to obtain vibration characteristic data; time synchronization and space registration are performed on the workpiece surface image data, the current density distribution data, the temperature field mapping data and the vibration characteristic data to obtain hanging plating line body state data; multi-band spectral acquisition is performed on the electroplating solution by using an ultraviolet-visible spectrum sensor, a near-infrared spectrum sensor and a Raman spectrum sensor to obtain electroplating solution multi-spectral data.

3. The full-automatic intelligent continuous hanging plating line body detection method according to claim 2, characterized in that, The joint feature extraction performed on the hanging plating line body state data and the electroplating solution multi-spectral data to obtain a feature vector representing the relationship between the hanging plating process and the electroplating solution composition comprises: electroplating defect recognition and defect feature parameter extraction are performed on the workpiece surface image data in the hanging plating line body state data to obtain an electroplating quality feature set; electrofield uniformity calculation is performed on the current density distribution data in the hanging plating line body state data to obtain an electroplating uniformity feature set; spectrum feature extraction is performed on the electroplating solution multi-spectral data to obtain an electroplating solution composition feature set; time sequence correlation analysis is performed on the electroplating quality feature set and the electroplating solution composition feature set to obtain quality-composition correlation features; space correlation analysis is performed on the electroplating uniformity feature set and the electroplating solution composition feature set, and a hot spot mapping algorithm is used to identify key influence areas to obtain uniformity-composition correlation features; the quality-composition correlation features and the uniformity-composition correlation features are fused to form a feature vector representing the relationship between the hanging plating process and the electroplating solution composition.

4. The full-automatic intelligent continuous hanging plating line body detection method according to claim 1, characterized in that, The association function between the hanging plating line body running state and the electroplating solution recovery efficiency is constructed based on the feature vector to obtain a mapping model of the hanging plating production quality and the recovery parameters, which comprises: nonlinear dimension reduction is performed on the feature vector to obtain hanging-plating-recovery feature dimension reduction data; a multiple-input multiple-output relationship network is constructed using the hanging-plating-recovery feature dimension reduction data, and a mapping relationship between the feature vector and the recovery efficiency is established through Bayesian sparse parameter polynomial regression to obtain a preliminary correlation function; a hanging-plating line body operating condition classification mechanism is introduced for the preliminary correlation function, a typical operating condition is identified through a clustering algorithm, and a target sub-model is trained for each type of operating condition to obtain an operating condition adaptive correlation function; a bidirectional mapping between electroplating quality and recovery parameters is constructed based on the operating condition adaptive correlation function, and an optimal recovery parameter is deduced from a target quality through a reverse reasoning algorithm to obtain a recovery parameter prediction model; a speed coordination response model of line speed and recovery rate is established in combination with the recovery parameter prediction model and real-time hanging-plating line speed; the speed coordination response model is subjected to liquid level balance constraints with an electroplating tank liquid level control strategy to obtain a mapping model of hanging-plating production quality and recovery parameters.

5. The fully automatic intelligent continuous hanging plating line body detection method according to claim 1, characterized in that, a variable genetic factor-based multi-objective optimization algorithm is executed according to the mapping model, and a hierarchical collaborative control strategy is created to generate linkage control parameters of hanging-plating line body production parameters and electroplating liquid recovery equipment, including: the mapping model is converted into a multi-objective optimization problem, and a target weight configuration of the multi-objective optimization problem is set based on a variable genetic factor-based multi-objective optimization algorithm; a hanging-plating line body control layer is constructed based on the target weight configuration, and a hanging-plating production control instruction is generated according to the hanging-plating line body control layer; an electroplating liquid recovery equipment control layer is constructed according to the mapping model, and a recovery equipment control instruction is calculated through the electroplating liquid recovery equipment control layer; a collaborative execution mechanism of the hanging-plating production control instruction and the recovery equipment control instruction is designed, and a parameter change transfer function and interlocking protection logic are established according to the collaborative execution mechanism to obtain a hierarchical collaborative control strategy; a global resource scheduling layer of the hanging-plating line body and the recovery equipment is constructed, and an overall optimization problem under resource constraints is solved according to the global resource scheduling layer to obtain an optimal allocation scheme; the hierarchical collaborative control strategy and the optimal allocation scheme are integrated, and control conversion logic of different production stages is managed through a state machine to obtain linkage control parameters of hanging-plating line body production parameters and electroplating liquid recovery equipment.

6. The full-automatic intelligent continuous hanging plating line body detection method according to claim 5, characterized in that, the mapping model is converted into a multi-objective optimization problem, and a target weight configuration of the multi-objective optimization problem is set based on a variable genetic factor-based multi-objective optimization algorithm, including: according to the mapping model, an electroplating quality index is converted into an absolute error function, a recovery efficiency index is converted into a relative efficiency function, and an energy consumption index is converted into a unit cost function to obtain a multi-objective function set; according to the mapping model, hanging-plating line body control parameters and electroplating liquid recovery equipment control parameters are uniformly encoded into a multi-dimensional vector, and physical constraints of each parameter are converted into variable boundary conditions to obtain a decision space; process constraint conditions are applied to the decision space, and a coupling constraint equation between electroplating quality and recovery efficiency and a balance constraint equation between control parameters are established to obtain a multi-objective optimization problem; mapping decision variables to binary or real number chromosomes based on the multi-objective optimization problem, and defining a fitness function to reflect the pros and cons of the chromosomes; introducing a genetic factor adjustment mechanism of a multi-objective optimization algorithm based on the multi-objective optimization problem and the fitness function, and dynamically calculating a crossover probability and a mutation probability based on the genetic factor adjustment mechanism to obtain adaptive genetic parameters; performing multi-objective evolutionary calculation using the adaptive genetic parameters to obtain a candidate weight configuration set, and selecting an optimal solution based on the decision preference function and the candidate weight configuration set to obtain a target weight configuration of the multi-objective optimization problem.

7. The fully automatic intelligent continuous hanging plating line body detection method according to claim 1, characterized in that, The dynamic coordination of the plating speed and the electroplating liquid treatment speed based on the linkage control parameters is executed to obtain a synchronization strategy of the line body operation and the recovery treatment, which includes: According to the linkage control parameters, the proportional relationship analysis of the plating line speed and the electroplating liquid flow is executed to obtain line speed-flow matching data; Using the line speed-flow matching data, a dynamic flow adjustment mechanism is obtained for rapid response to changes in the plating line speed, and a smooth transition scheme for the workpiece switching process of the plating line body is obtained for transition period control; An adaptive mapping relationship between the component concentration and the treatment speed is established in combination with the real-time monitoring results of the electroplating liquid composition to obtain a concentration adaptive treatment scheme; A recovery equipment response analysis is performed for planned and unplanned shutdown of the plating line body to obtain a shutdown process treatment scheme; The dynamic flow adjustment mechanism, the smooth transition scheme, the concentration adaptive treatment scheme, and the shutdown process treatment scheme are integrated to obtain a synchronization strategy of the line body operation and the recovery treatment.

8. A full-automatic intelligent continuous hanging plating line body detection device, characterized in that, The full-automatic intelligent continuous plating line body detection device for executing the full-automatic intelligent continuous plating line body detection method as claimed in any one of claims 1-7, comprising: A data acquisition module for acquiring real-time multi-modal data of the plating line body operation process to obtain plating line body state data and electroplating liquid multi-spectral data; A feature extraction module for jointly extracting features from the plating line body state data and the electroplating liquid multi-spectral data to obtain a feature vector representing the relationship between the plating process and the electroplating liquid composition; A construction module for constructing a correlation function between the plating line body operation state and the electroplating liquid recovery efficiency based on the feature vector to obtain a mapping model of the plating production quality and the recovery parameters; A multi-objective optimization module for executing a multi-objective optimization algorithm based on variable genetic factors based on the mapping model and creating a hierarchical collaborative control strategy to generate linkage control parameters of the plating line body production parameters and the electroplating liquid recovery equipment; A dynamic coordination module for executing dynamic coordination of the plating speed and the electroplating liquid treatment speed based on the linkage control parameters to obtain a synchronization strategy of the line body operation and the recovery treatment.

Images