Online coating thickness x-ray real-time feedback regulation method and system for electroplating process

Abstract

The application provides an online coating thickness X-ray real-time feedback regulation method and system for an electroplating process. The application belongs to the technical field of precision manufacturing and process control. The method comprises the following steps: dividing a micro-area coating thickness trace measurement area of a moving workpiece, and generating dynamic micro-area measurement grid data; deploying an X-ray fluorescence scanning device according to the dynamic micro-area measurement grid data, and constructing an online trace measurement network; the online coating thickness X-ray real-time feedback regulation method for the electroplating process improves the coating thickness control precision, realizes accurate control of ±0.05 microns, enhances the coating uniformity, controls the uniformity to be less than or equal to ±3%, and meets the strict requirements in high-end fields.

Description

Technical Field

[0001] This invention proposes an online X-ray real-time feedback control method and system for coating thickness in electroplating processes, which belongs to the interdisciplinary field of precision manufacturing and process control. Background Technology

[0002] In the field of electroplating, precise control of coating thickness is crucial for ensuring product performance and quality. However, current electroplating processes face numerous challenges in coating thickness control that urgently need to be addressed.

[0003] Traditional methods primarily rely on offline sampling inspection, such as coulometric methods or metallographic sectioning. These methods can only detect defects after electroplating is complete and cannot intervene in the ongoing deposition process. Once an entire batch exceeds tolerances, the losses are difficult to recover. Although some online measurement systems have introduced beta-ray or eddy current thickness measurement, their spatial resolution is low, greater than 5 mm. 2 It is difficult to capture thickness changes in minute areas and cannot distinguish multi-layer structures. The measurement results are easily affected by the substrate, which greatly reduces the accuracy.

[0004] Regarding the control mechanism, even with online data, adjustments to current and time still rely on manual experience, lacking deep coupling with electrochemical mechanisms. This makes adaptive feedback impossible and hinders the guarantee of coating thickness uniformity and stability. More importantly, existing technologies do not treat coating growth as a continuous trace formation process, neglecting the impact of local anomalies such as edge effects, masking areas, and current distribution distortions on coating thickness.

[0005] Therefore, existing electroplating processes cannot meet the stringent requirements of high-end electronics, automotive, and aerospace industries for coating thickness control accuracy of ±0.05μm and uniformity of ≤±3%. Developing a novel online X-ray real-time feedback control method for coating thickness is urgently needed. Summary of the Invention

[0006] This invention provides a method and system for real-time X-ray feedback control of online coating thickness in electroplating processes, to solve the problems mentioned in the background section above: The present invention proposes an online X-ray real-time feedback control method for coating thickness in electroplating processes, the method comprising: S1. Divide the moving workpiece into micro-area coating thickness trace measurement areas and generate dynamic micro-area measurement grid data; deploy an X-ray fluorescence scanning device based on the dynamic micro-area measurement grid data to construct an online trace measurement network; S2. Based on the online trace measurement network, millisecond-level synchronous data acquisition is performed to generate a time-series dataset of coating thickness traces and workpiece motion trajectory data; the time-series dataset of coating thickness traces is processed to suppress substrate interference and separate multilayer structure signals to generate pure dynamic trace data of coating thickness. S3. Based on the dynamic trace data of pure coating thickness, the electrochemical deposition model is calibrated online to obtain local current density distribution data and coating growth rate field data respectively; nonlinear coupling analysis is performed on the coating growth rate field data through local current density distribution data to generate micro-area thickness deviation prediction data. S4. Use the workpiece motion trajectory data to perform spatial coordinate mapping processing on the micro-area thickness deviation prediction data to generate a full workpiece surface thickness deviation heatmap; use the full workpiece surface thickness deviation heatmap to identify edge effects and shading areas to generate local abnormal control area data; perform millisecond-level electroplating parameter closed-loop correction on the local abnormal control area data to generate a real-time control instruction set. S5. Perform multi-actuator collaborative control based on the real-time control instruction set to generate a dynamic electroplating process parameter field; calculate the coating thickness stability index based on the dynamic electroplating process parameter field to generate comprehensive electroplating quality assessment data; trigger a graded early warning mechanism based on the comprehensive electroplating quality assessment data to generate intelligent control feedback data for the electroplating process.

[0007] The present invention proposes a system for implementing an online coating thickness real-time feedback control method using X-rays in the electroplating process described above, the system comprising: Network construction module: Divides the moving workpiece into micro-area coating thickness trace measurement regions, generating dynamic micro-area measurement grid data; deploys an X-ray fluorescence scanning device based on the dynamic micro-area measurement grid data to construct a high-resolution (≤1mm) network. 2 Online trace measurement network; Signal processing module: Based on the online trace measurement network, it performs millisecond-level synchronous data acquisition to generate a time-series dataset of coating thickness traces and workpiece motion trajectory data; it performs substrate interference suppression and multi-layer structure signal separation processing on the time-series dataset of coating thickness traces to generate clean dynamic trace data of coating thickness. Coupling analysis module: Based on the dynamic trace data of pure coating thickness, the electrochemical deposition model is calibrated online to obtain local current density distribution data and coating growth rate field data respectively; Nonlinear coupling analysis is performed on the coating growth rate field data through local current density distribution data to generate micro-area thickness deviation prediction data; Command control module: It uses workpiece motion trajectory data to perform spatial coordinate mapping processing on micro-area thickness deviation prediction data to generate a full workpiece surface thickness deviation heatmap; it uses the full workpiece surface thickness deviation heatmap to identify edge effects and shading areas, generating local abnormal control area data; it performs millisecond-level electroplating parameter closed-loop correction on the local abnormal control area data to generate a real-time control command set; The graded early warning module performs collaborative control of multiple actuators based on the real-time control instruction set to generate a dynamic electroplating process parameter field; it calculates the coating thickness stability index based on the dynamic electroplating process parameter field to generate comprehensive electroplating quality assessment data; and it triggers a graded early warning mechanism based on the comprehensive electroplating quality assessment data to generate intelligent control feedback data for the electroplating process.

[0008] The beneficial effects of this invention are as follows: The online X-ray real-time feedback control method for electroplating thickness in this invention improves the accuracy of coating thickness control, achieving precise control of ±0.05μm, while enhancing coating uniformity, controlling it to ≤±3%, meeting the stringent requirements of high-end fields. This method reduces reliance on manual intervention, overcoming the limitations of traditional manual experience-based adjustments through an adaptive feedback mechanism deeply coupled with electrochemical mechanisms, thus reducing human error. It also reduces scrap rates, corrects electroplating parameters in milliseconds, promptly handles local anomalies, and avoids batch-wide deviations. It can acquire micro-area coating thickness information in real time through a high-resolution online trace measurement network during high-speed production of moving workpieces, and can also perform multi-actuator collaborative control based on dynamic data to achieve efficient and stable production. Furthermore, it avoids the lag of traditional offline sampling inspection, allows real-time intervention in the deposition process to prevent further losses, and significantly improves the intelligence level and production efficiency of the electroplating process. Attached Figure Description

[0009] Figure 1 This is a diagram illustrating the steps of the method described in this invention; Figure 2 This is a system module diagram of the present invention. Detailed Implementation

[0010] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0011] One embodiment of the present invention, such as Figure 1 As shown, an online X-ray real-time feedback control method for coating thickness in electroplating processes is provided, the method comprising: S1. Divide the moving workpiece into micro-area coating thickness trace measurement areas to generate dynamic micro-area measurement grid data; deploy an X-ray fluorescence scanning device based on the dynamic micro-area measurement grid data to construct a high-resolution (≤1mm) measurement grid. 2 Online trace measurement network; S2. Based on the online trace measurement network, millisecond-level synchronous data acquisition is performed to generate a time-series dataset of coating thickness traces and workpiece motion trajectory data; the time-series dataset of coating thickness traces is processed to suppress substrate interference and separate multilayer structure signals to generate pure dynamic trace data of coating thickness. S3. Based on the dynamic trace data of pure coating thickness, the electrochemical deposition model is calibrated online to obtain local current density distribution data and coating growth rate field data respectively; nonlinear coupling analysis is performed on the coating growth rate field data through local current density distribution data to generate micro-area thickness deviation prediction data. S4. Use the workpiece motion trajectory data to perform spatial coordinate mapping processing on the micro-area thickness deviation prediction data to generate a full workpiece surface thickness deviation heatmap; use the full workpiece surface thickness deviation heatmap to identify edge effects and shading areas to generate local abnormal control area data; perform millisecond-level electroplating parameter (current / time) closed-loop correction on the local abnormal control area data to generate a real-time control instruction set. S5. Perform multi-actuator collaborative control based on the real-time control instruction set to generate a dynamic electroplating process parameter field; calculate the coating thickness stability index based on the dynamic electroplating process parameter field to generate comprehensive electroplating quality assessment data; trigger a graded early warning mechanism based on the comprehensive electroplating quality assessment data to generate intelligent control feedback data for the electroplating process.

[0012] The working principle and effects of the above technical solution are as follows: Online X-ray scanning and precise micro-area segmentation improve the resolution and real-time performance of coating thickness measurement, enhancing the targeted nature of process control. Substrate interference suppression and multi-layer signal separation processing reduce coating thickness data deviations, preventing control decision errors caused by signal interference and thus reducing the yield of defective products. Online correction of the electrochemical deposition model and nonlinear coupling analysis improve the accuracy of micro-area thickness deviation prediction, enhance the precision of parameter control, and reduce resource waste caused by blind adjustments to electroplating parameters. Millisecond-level closed-loop correction and multi-mechanism collaborative control accelerate the response speed of control in abnormal areas, avoiding coating unevenness and thickness deviation problems caused by lagging control, and preventing the accumulation of quality hazards. Dynamic process parameter fields and hierarchical early warning linkage ensure coating thickness stability and overall electroplating quality, while also improving the continuity of process operation, reducing downtime for maintenance, lowering production and operating costs, and making the electroplating process more adaptable to the needs of efficient and precise production.

[0013] In one embodiment of the present invention, S1 includes: S11. Collect the external dimensions, movement speed, and basic parameters of the electroplating process of the moving workpiece to generate a comprehensive feature dataset of the moving workpiece. S12. Based on the comprehensive feature dataset of the moving workpiece, the micro-area coating thickness trace measurement area of ​​the moving workpiece is divided, and dynamic micro-area measurement grid data is generated. S13. Based on the dynamic micro-area measurement grid data, plan the deployment locations of the X-ray fluorescence scanning device and generate a dataset of device deployment location distribution. S14. Complete the installation and commissioning of the X-ray fluorescence scanning device according to the device deployment point distribution dataset, and generate the device linkage operation dataset. S15. Based on the device linkage operation dataset, start the collaborative working program of each scanning device to build a high-resolution online trace measurement network.

[0014] The working principle and effects of the above technical solution are as follows: By comprehensively collecting multi-dimensional parameters of the moving workpiece to generate a comprehensive feature dataset, sufficient data support can be provided for micro-area division, avoiding unreasonable measurement area division due to missing parameters and reducing subsequent scanning deviations. Based on this dataset, micro-area division is carried out to generate a dynamic mesh, improving the adaptability of the measurement area and enhancing the adaptability to complex shapes and motion states of the workpiece, avoiding scanning blind spots or redundant measurements caused by fixed meshes. The scanning device points are planned according to the mesh data, optimizing the rationality of the device layout and reducing resource consumption caused by idle equipment or overly dense layout. Installation and debugging are completed according to the points to generate a linkage dataset, ensuring the consistency of operation of each device and avoiding the impact of single device failure or poor coordination on measurement efficiency. The collaborative program is launched to build a high-resolution network, which not only improves the accuracy of coating trace measurement but also enhances the stability of online measurement, laying a solid foundation for subsequent data acquisition and preventing the overall control process from lagging due to measurement network defects.

[0015] In one embodiment of the present invention, S12 includes: S121. Import the comprehensive feature dataset of moving workpieces, extract the shape contour features, motion speed features and electroplating process related features of the moving workpieces, and generate a subset of core features of the moving workpieces. S122. Based on the core feature subset of the moving workpiece, the surface of the moving workpiece is pre-divided into uniformities to generate an initial micro-region division set for the surface of the moving workpiece. S123. Combine the motion rate characteristics of the moving workpiece to perform spatiotemporal dimension adaptation and adjustment on the initial micro-region division set to generate a set of spatiotemporal adjustment factors for the micro-region division. S124. Optimize the boundaries of the initial micro-region division set by using the micro-region division spatiotemporal adjustment factor set to generate the micro-region division correction set of the moving workpiece surface. S125. Perform mesh cell normalization on the micro-area division correction region set of the moving workpiece surface to generate dynamic micro-area measurement mesh data.

[0016] The working principle and effects of the above technical solution are as follows: By extracting a subset of the core features of the workpiece and removing redundant parameters, the relevance of the feature data is improved, irrelevant information is avoided from interfering with the accuracy of micro-area division, and the workload of subsequent adjustments is reduced. Uniform pre-division is performed based on the core features, making the initial micro-area layout more closely match the shape of the workpiece, enhancing the basic rationality of the division, and preventing local areas from being divided too sparsely or too densely. Spatiotemporal dimension adjustment is combined with the movement rate to adapt the micro-area division to the dynamic changes of the workpiece, avoiding measurement deviations caused by the disconnect between fixed area division and workpiece movement. Adjustment factors are used to optimize the area boundaries, correcting irregularities in the initial division, reducing scanning blind spots and area overlap, and improving the rigor of micro-area division. The corrected area is then subjected to grid normalization processing to unify the grid unit specifications, ensuring accurate docking of subsequent scanning devices and improving the consistency of data acquisition. This avoids measurement data distortion caused by chaotic grid specifications, laying a solid foundation for high-resolution online measurement and improving the smoothness of the overall process.

[0017] In one embodiment of the present invention, step S123 includes: Extract the motion rate features from the core feature subset of the moving workpiece, break down the rate fluctuation pattern, instantaneous rate peak and rate stability index, and generate a rate feature sequence set; The set of rate feature sequences is matched one by one with the initial set of micro-region divisions. The spatiotemporal correlation analysis of micro-region location and rate change is carried out to generate a set of regional spatiotemporal correlation coefficients. Based on the set of regional spatiotemporal correlation coefficients, the boundary range and unit density of the initial micro-region division set are initially adapted and adjusted to generate a spatiotemporally adapted preliminary adjusted region set. Verify the rationality of the matching between the initial spatiotemporal adaptation adjustment region set and the motion rate fluctuation, investigate issues such as region overlap and scanning blind spots, and generate adaptation deviation correction factors; By optimizing the spatiotemporal adaptation region set through the adaptation deviation correction factor, a spatiotemporal adjustment factor set for micro-region division is generated.

[0018] The working principle and effects of the above technical solution are as follows: By decomposing the fluctuation patterns, peak values, and stability indicators of the motion rate to generate a sequence set, the dimension of rate characteristics is refined, improving the control over the workpiece's motion state, avoiding adaptation deviations caused by a single rate parameter, and reducing the blindness of subsequent adjustments. Correlation analysis is performed one-to-one between the rate sequence and the initial micro-regions to clarify the spatiotemporal relationship between the micro-region position and rate changes, enhancing the targeted nature of the division and adjustment, and preventing the micro-region layout from becoming disconnected from the workpiece's motion rhythm. Based on the correlation coefficient, the micro-region boundaries and density are initially adjusted to better adapt to dynamic rate changes, improving the basic rationality of spatiotemporal adaptation and reducing measurement loopholes caused by fixed layouts. The initial adjustment results are verified, and overlap and blind zone issues are investigated to promptly identify adaptation defects. Correction factors are generated to compensate for previous deviations, preventing problems from affecting subsequent processes. The set of adjustment regions is optimized through correction factors to generate a precise set of spatiotemporal adjustment factors, which not only improves the fit between micro-region division and workpiece motion but also provides a reliable basis for subsequent boundary optimization, preventing scanning data distortion due to insufficient adaptation, and further ensuring the accuracy and smoothness of the overall measurement process.

[0019] In one embodiment of the present invention, S2 includes: S21. Start the high-resolution online trace measurement network to carry out millisecond-level synchronous data acquisition and generate a time-series dataset of coating thickness traces and workpiece motion trajectory data; S22. Perform high-frequency noise filtering on the coating thickness trace time series dataset to generate a noise-reduced coating thickness time series dataset; extract the substrate signal features from the noise-reduced coating thickness time series dataset to generate a substrate signal feature benchmark dataset. S23. Based on the substrate signal feature benchmark dataset, perform substrate interference suppression on the noise reduction coating thickness time series dataset to generate substrate interference-free coating time series data. S24. Perform multi-layer structure signal separation processing on the substrate interference removal coating timing data to generate pure coating thickness dynamic trace data.

[0020] The working principle and effects of the above technical solution are as follows: A high-resolution network is activated to perform millisecond-level synchronous acquisition, accelerating the acquisition efficiency of coating thickness and workpiece trajectory data, ensuring data timeliness, avoiding incomplete dynamic trace recording due to acquisition lag, and reducing the workload of subsequent data completion. High-frequency noise is filtered on the time-series dataset to remove irrelevant interference signals, improving data purity and preventing noise from affecting the accuracy of coating thickness feature extraction. Substrate signal features are extracted to generate a benchmark dataset, providing a precise reference for interference suppression, enhancing the targeting of substrate signal removal, preventing substrate signals from obscuring the true coating data, and reducing thickness detection deviation. Substrate interference is suppressed based on the benchmark dataset, purifying the coating time-series data, making the coating's own characteristics more prominent, and creating favorable conditions for multi-layer signal separation. Multi-layer signal separation is performed on the interference-removed data, splitting the superimposed signals of each coating layer to generate pure dynamic trace data. This ensures the authenticity and accuracy of the coating thickness data and provides reliable input for subsequent model calibration, avoiding control decision deviations caused by signal superposition, and further strengthening the data analysis foundation of the overall feedback control process.

[0021] In one embodiment of the present invention, S24 includes: Analyze the timing data of the substrate-interference-free coating, separate the signal superposition components between each coating layer, and generate a multi-layer signal superposition feature set; Based on the feature set of multi-layer signal superposition, the differences in signal frequency and amplitude of different coating layers are distinguished, and a threshold set for inter-layer signal distinction is generated. Based on the inter-layer signal differentiation threshold set, the signal is stripped layer by layer from the multi-layer signal superposition feature set to generate a single-level signal dataset. Verify the integrity of a single-level signal dataset, filter out residual interference signals between levels, and generate a purified hierarchical signal dataset. By integrating signal purification datasets from various levels, the dynamic variation pattern of coating thickness is restored, and clean coating thickness dynamic trace data is generated.

[0022] The working principle and effects of the above technical solution are as follows: It analyzes and de-interfering coating time-series data and separates interlayer superimposed signals, clarifying the superposition logic of signals from each coating layer, reducing the problem of signal masking between different coating layers, and avoiding thickness data distortion caused by direct analysis. Based on superposition characteristics, it distinguishes the signal frequency and amplitude differences of different coating layers, generating accurate interlayer discrimination thresholds, improving the targeting of interlayer signal separation, and preventing incomplete signal stripping caused by threshold ambiguity. It strips signals layer by layer according to the threshold, separating single-layer datasets, enhancing the independence of each coating layer signal, reducing interlayer cross-interference, and creating conditions for separate analysis of thickness changes in each layer. It verifies the integrity of the single-layer dataset and filters residual interference, further purifying signal quality, avoiding residual impurities from affecting data accuracy, and reducing the risk of errors in subsequent analysis. It integrates the purified data from each layer, restoring the dynamic change law of coating thickness, and generating pure trace data, which can accurately reflect the true growth state of each coating layer and provide high-quality input for subsequent electrochemical model calibration, avoiding model calibration failure due to data deviation, and effectively ensuring the accuracy of overall feedback control.

[0023] In one embodiment of the present invention, S3 includes: S31. Import the dynamic trace data of the pure coating thickness and generate the input dataset for the electrochemical deposition model calibration. S32. Based on the input dataset of the electrochemical deposition model, iteratively update the parameters of the electrochemical deposition model to generate the corrected electrochemical deposition model. S33. Calculate the local current density distribution data using the corrected electrochemical deposition model, and generate a local current density distribution dataset. S34. Calculate the coating growth rate field data based on the local current density distribution dataset, and generate the coating growth rate field dataset. S35. Perform nonlinear coupling analysis on the local current density distribution dataset and the coating growth rate field dataset to generate micro-area thickness deviation prediction data.

[0024] The working principle and effects of the above technical solution are as follows: Importing pure coating thickness dynamic trace data to generate a model calibration input set ensures the accuracy of the input data, avoids impurities causing the model calibration to deviate from reality, and reduces subsequent calculation errors. Based on the input set, iteratively updating the electrochemical deposition model parameters makes the model more closely match real-time electroplating conditions, improves model adaptability, enhances the reliability of data calculation, and prevents result deviations caused by outdated model parameters. Using the calibrated model, calculating the local current density distribution refines the spatial distribution characteristics of the current density, improves the accuracy of data calculation, and reduces current density judgment errors caused by model bias. Based on the current density data, calculating the coating growth rate field clearly presents the coating growth rhythm in different regions, enhances the visualization and controllability of the rate distribution, and avoids misjudgments of local growth rates affecting regulation. Nonlinear coupling analysis is performed on the two types of datasets to explore the intrinsic relationship between current density and growth rate, improving the accuracy of micro-area thickness deviation prediction and avoiding prediction loopholes caused by single-dimensional analysis. This provides accurate basis for subsequent abnormal area location and can predict thickness deviation trends in advance, preventing coating quality problems caused by the expansion of deviations, and effectively improving the precise control capability of the electroplating process.

[0025] In one embodiment of the present invention, S33 includes: Extract the core parameters of the corrected electrochemical deposition model, and decompose the electrode reaction coefficient, electrolyte conductivity and interfacial impedance parameters to generate the core parameter set of the model. The core parameter set of the correlation model and the calibration input dataset of the electrochemical deposition model are used to extract the correlation features of micro-area coating growth and generate the input set for current density calculation. Based on the current density calculation input set, the point-by-point current density calculation of the micro-region is carried out through the corrected model to generate a preliminary local current density dataset. By comparing the correlation between the preliminary dataset and the dynamic trace data of coating thickness, abnormal calculation results are filtered out, and a current density correction dataset is generated. The spatial coordinates and numerical accuracy of the dataset are regularized and corrected to generate a local current density distribution dataset.

[0026] The working principle and effects of the above technical solution are as follows: The core parameters of the corrected model are extracted and key indicators are broken down to generate a precise set of core model parameters. This refines the parameter dimensions, improves the targeting of parameter calls, avoids computational bias caused by mixed parameters, and reduces the amount of unnecessary work in subsequent calculations. The core parameter set is correlated with the model correction input dataset to extract micro-area coating growth correlation features, generating a current density calculation input set. This enhances the fit between the calculation input and actual working conditions, preventing calculation distortion caused by input data disconnect. Based on the calculation input set, point-by-point current density calculations are performed in the micro-area to generate a preliminary local current density dataset. This refines the spatial distribution details of the current density, improves the detail of the calculation results, and avoids micro-area data loss caused by overall coarse calculations. The preliminary dataset is compared with dynamic trace data of coating thickness to filter out abnormal calculation results, generating a corrected current density dataset. This improves data reliability and prevents abnormal data from interfering with subsequent analysis and judgment. By standardizing and correcting the spatial coordinates and numerical accuracy of the dataset, a local current density distribution dataset is generated. This unifies the data standards, ensuring the accuracy of subsequent coating growth rate field calculations and providing high-quality data support for micro-area thickness deviation prediction. It avoids analytical errors caused by inconsistent data specifications and effectively improves the quality of the data foundation for electroplating process control.

[0027] In one embodiment of the present invention, step S4 includes: S41. Import micro-area thickness deviation prediction data and workpiece motion trajectory data to generate spatial coordinate mapping analysis input dataset; S42. Perform full-dimensional coordinate matching on the input dataset based on spatial coordinate mapping analysis to generate a coordinate matching dataset; S43. Generate a thermal map of the thickness deviation of the entire workpiece surface using the coordinate matching dataset; S44. Perform edge effect and shading area feature identification on the thermal map of thickness deviation of the entire workpiece surface to generate local anomaly control area data; S45. Based on the data of the local abnormal control area, perform millisecond-level closed-loop correction of electroplating parameters to generate a real-time control instruction set.

[0028] The working principle and effects of the above technical solution are as follows: An input set is generated by importing micro-area thickness deviation prediction data and workpiece motion trajectory data, integrating these two types of core data to avoid inaccurate deviation positioning caused by isolated data and reduce the blindness of subsequent analysis. Based on the input set, full-dimensional coordinate matching is performed to generate a coordinate matching dataset, improving the correspondence accuracy between deviation data and the actual position of the workpiece and preventing misjudgment of abnormal areas caused by coordinate misalignment. A thickness deviation heatmap is generated using the coordinate matching dataset, transforming abstract data into intuitive images, enhancing the visualization of deviation distribution, and reducing the tedious process and errors of manual analysis. Edge effect and occlusion area feature recognition is performed on the heatmap to generate local abnormal control area data, accurately locating problem areas, improving the targeting of control, and avoiding resource waste caused by indiscriminate control. Based on the abnormal area data, millisecond-level closed-loop correction of electroplating parameters is performed to generate a real-time control instruction set, accelerating the response speed of parameter adjustments and avoiding the expansion of coating thickness deviation caused by control lag. This not only ensures the uniformity of coating thickness across all workpieces but also enhances the dynamic control capability of the electroplating process, effectively reducing the yield of defective products and stabilizing the overall stability of electroplating quality.

[0029] In one embodiment of the present invention, step S5 includes: S51. Import the real-time control instruction set and generate a multi-actuator collaborative control input dataset; S52. Based on the multi-actuator collaborative control input dataset, drive the linkage operation of each actuator to generate a dynamic electroplating process parameter field; S53. Collect real-time monitoring data of coating thickness under dynamic electroplating process parameter field, and generate input dataset for stability index calculation; S54. Calculate the coating thickness stability index on the input dataset for stability index calculation, and generate comprehensive electroplating quality assessment data. S55. Based on comprehensive electroplating quality assessment data, a graded early warning mechanism is triggered to generate intelligent control feedback data for the electroplating process.

[0030] The working principle and effects of the above technical solution are as follows: Importing a real-time control instruction set generates a multi-actuator collaborative control input set, standardizing instruction transmission logic, avoiding disconnection or inconsistent responses among mechanisms, and reducing linkage operation failures. Based on the input set, the actuators are driven to operate in conjunction, generating a dynamic electroplating process parameter field. This allows process parameters to be adjusted in real-time according to operating conditions, enhancing parameter flexibility and preventing quality fluctuations caused by fixed parameters being unable to cope with complex electroplating scenarios. Real-time monitoring data of plating thickness under the dynamic parameter field is collected to generate a stability index calculation input set, ensuring that the evaluation data closely matches real-time operating conditions and avoiding quality misjudgments caused by calculations based on lagging data. The stability index is calculated on the input set to generate comprehensive electroplating quality evaluation data, refining the quality evaluation dimensions, improving the accuracy of plating quality control, and reducing the omission of hidden thickness deviations. Based on the evaluation data, a graded early warning mechanism is triggered, generating intelligent control feedback data. This enables closed-loop optimization and adjustment of the process, and also avoids quality risks in advance, preventing small deviations from gradually expanding and causing batches of unqualified products. It effectively reduces rework costs and production losses, while enhancing the automation control capabilities of the electroplating process and consolidating the consistency and reliability of overall production quality.

[0031] In one embodiment of the present invention, S52 includes: Analyze the input dataset for multi-actuator coordinated control, break down the action commands, operating thresholds and linkage timing of each actuator, and generate a set of actuator control parameters; The mechanism control parameter set is assigned to the corresponding actuator one by one, the mechanism initialization program is started, and the mechanism start-up status dataset is generated. Based on the mechanism startup status dataset, the actuators are synchronously driven to run in the linkage sequence, and the mechanism operation parameters and process feedback data are collected in real time to generate a real-time dataset of linkage operation. By comparing the real-time dataset of the linkage operation with the set of mechanism control parameters, parameter deviations and mechanism action lag issues are corrected, and a set of process parameter corrections is generated. By integrating the process parameter correction set with the electroplating conditions of each region, a set of process parameters that change dynamically in time and space is formed, generating a dynamic electroplating process parameter field.

[0032] The working principle and effects of the above technical solution are as follows: The collaborative control input dataset is parsed and the mechanism action commands, thresholds, and linkage timing are broken down to generate a standardized set of mechanism control parameters. This clarifies the operating standards of each mechanism, avoids action confusion caused by mixed commands, and reduces mechanism linkage failures. The parameter set is assigned to the corresponding mechanism one by one and initialized, generating a startup status dataset. Anomalies in mechanism startup are promptly identified to prevent individual mechanism failures from affecting the overall linkage effect and improve operational stability. Based on the startup status dataset, the mechanism is synchronously driven to operate according to the timing sequence. Operating parameters and process feedback data are collected in real time to generate a real-time linkage operation dataset, accurately capturing dynamic changes in operating conditions and preventing action lag or parameter deviations from going undetected. The real-time dataset is compared with the control parameter set to correct deviations and lag issues, generating a process parameter correction set to improve parameter adaptability and reduce coating quality problems caused by accumulated deviations. By integrating the correction set with the electroplating conditions of each region to generate a dynamic process parameter field, the parameters can be flexibly adjusted according to the spatiotemporal changes of the working conditions, while ensuring the consistency of the process in each region. This avoids the drawbacks of fixed parameters being unable to adapt to complex working conditions, effectively enhancing the dynamic control capability of the electroplating process and stabilizing the stability of the coating quality.

[0033] One embodiment of the present invention, such as Figure 2 As shown, a system for implementing the online X-ray real-time feedback control method for electroplating thickness in the above-described electroplating process is provided, the system comprising: Network construction module: Divides the moving workpiece into micro-area coating thickness trace measurement regions, generating dynamic micro-area measurement grid data; deploys an X-ray fluorescence scanning device based on the dynamic micro-area measurement grid data to construct a high-resolution (≤1mm) network. 2 Online trace measurement network; Signal processing module: Based on the online trace measurement network, it performs millisecond-level synchronous data acquisition to generate a time-series dataset of coating thickness traces and workpiece motion trajectory data; it performs substrate interference suppression and multi-layer structure signal separation processing on the time-series dataset of coating thickness traces to generate clean dynamic trace data of coating thickness. Coupling analysis module: Based on the dynamic trace data of pure coating thickness, the electrochemical deposition model is calibrated online to obtain local current density distribution data and coating growth rate field data respectively; Nonlinear coupling analysis is performed on the coating growth rate field data through local current density distribution data to generate micro-area thickness deviation prediction data; Command control module: It uses workpiece motion trajectory data to perform spatial coordinate mapping processing on micro-area thickness deviation prediction data to generate a full workpiece surface thickness deviation heat map; it uses the full workpiece surface thickness deviation heat map to identify edge effects and shading areas, generating local abnormal control area data; it performs millisecond-level electroplating parameter (current / time) closed-loop correction on the local abnormal control area data to generate a real-time control command set; The graded early warning module performs collaborative control of multiple actuators based on the real-time control instruction set to generate a dynamic electroplating process parameter field; it calculates the coating thickness stability index based on the dynamic electroplating process parameter field to generate comprehensive electroplating quality assessment data; and it triggers a graded early warning mechanism based on the comprehensive electroplating quality assessment data to generate intelligent control feedback data for the electroplating process.

[0034] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for real-time X-ray feedback control of online coating thickness in electroplating processes, characterized in that, The method includes: S1. Divide the moving workpiece into micro-area coating thickness trace measurement areas and generate dynamic micro-area measurement grid data; deploy an X-ray fluorescence scanning device based on the dynamic micro-area measurement grid data to construct an online trace measurement network; S2. Based on the online trace measurement network, millisecond-level synchronous data acquisition is performed to generate a time-series dataset of coating thickness traces and workpiece motion trajectory data; the time-series dataset of coating thickness traces is processed to suppress substrate interference and separate multilayer structure signals to generate pure dynamic trace data of coating thickness. S3. Based on the dynamic trace data of pure coating thickness, the electrochemical deposition model is calibrated online to obtain local current density distribution data and coating growth rate field data respectively; nonlinear coupling analysis is performed on the coating growth rate field data through local current density distribution data to generate micro-area thickness deviation prediction data. S4. Use the workpiece motion trajectory data to perform spatial coordinate mapping processing on the micro-area thickness deviation prediction data to generate a full workpiece surface thickness deviation heatmap; use the full workpiece surface thickness deviation heatmap to identify edge effects and shading areas to generate local abnormal control area data; perform millisecond-level electroplating parameter closed-loop correction on the local abnormal control area data to generate a real-time control instruction set. S5. Perform multi-actuator collaborative control based on the real-time control instruction set to generate a dynamic electroplating process parameter field; calculate the coating thickness stability index based on the dynamic electroplating process parameter field to generate comprehensive electroplating quality assessment data; trigger a graded early warning mechanism based on the comprehensive electroplating quality assessment data to generate intelligent control feedback data for the electroplating process.

2. The online X-ray real-time feedback control method for coating thickness in electroplating processes according to claim 1, characterized in that, S1 includes: S11. Collect the external dimensions, movement speed, and basic parameters of the electroplating process of the moving workpiece to generate a comprehensive feature dataset of the moving workpiece. S12. Based on the comprehensive feature dataset of the moving workpiece, the micro-area coating thickness trace measurement area of ​​the moving workpiece is divided, and dynamic micro-area measurement grid data is generated. S13. Based on the dynamic micro-area measurement grid data, plan the deployment locations of the X-ray fluorescence scanning device and generate a dataset of device deployment location distribution. S14. Complete the installation and commissioning of the X-ray fluorescence scanning device according to the device deployment point distribution dataset, and generate the device linkage operation dataset. S15. Based on the device linkage operation dataset, start the collaborative working program of each scanning device to build a high-resolution online trace measurement network.

3. The online X-ray real-time feedback control method for coating thickness in electroplating processes according to claim 2, characterized in that, S12 includes: S121. Import the comprehensive feature dataset of moving workpieces, extract the shape contour features, motion speed features and electroplating process related features of the moving workpieces, and generate a subset of core features of the moving workpieces. S122. Based on the core feature subset of the moving workpiece, the surface of the moving workpiece is pre-divided into uniformities to generate an initial micro-region division set for the surface of the moving workpiece. S123. Combine the motion rate characteristics of the moving workpiece to perform spatiotemporal dimension adaptation and adjustment on the initial micro-region division set to generate a set of spatiotemporal adjustment factors for the micro-region division. S124. Optimize the boundaries of the initial micro-region division set by using the micro-region division spatiotemporal adjustment factor set to generate the micro-region division correction set of the moving workpiece surface. S125. Perform mesh cell normalization on the micro-area division correction region set of the moving workpiece surface to generate dynamic micro-area measurement mesh data.

4. The online X-ray real-time feedback control method for coating thickness in electroplating processes according to claim 3, characterized in that, S123 includes: Extract the motion rate features from the core feature subset of the moving workpiece, break down the rate fluctuation pattern, instantaneous rate peak and rate stability index, and generate a rate feature sequence set; The set of rate feature sequences is matched one by one with the initial set of micro-region divisions. The spatiotemporal correlation analysis of micro-region location and rate change is carried out to generate a set of regional spatiotemporal correlation coefficients. Based on the set of regional spatiotemporal correlation coefficients, the boundary range and unit density of the initial micro-region division set are initially adapted and adjusted to generate a spatiotemporally adapted preliminary adjusted region set. Verify the rationality of the matching between the initial spatiotemporal adaptation adjustment region set and the motion rate fluctuation, investigate issues such as region overlap and scanning blind spots, and generate adaptation deviation correction factors; By optimizing the spatiotemporal adaptation region set through the adaptation deviation correction factor, a spatiotemporal adjustment factor set for micro-region division is generated.

5. The online X-ray real-time feedback control method for coating thickness in electroplating processes according to claim 1, characterized in that, S2 includes: S21. Start the high-resolution online trace measurement network to carry out millisecond-level synchronous data acquisition and generate a time-series dataset of coating thickness traces and workpiece motion trajectory data; S22. Perform high-frequency noise filtering on the coating thickness trace time series dataset to generate a noise-reduced coating thickness time series dataset; extract the substrate signal features from the noise-reduced coating thickness time series dataset to generate a substrate signal feature benchmark dataset. S23. Based on the substrate signal feature benchmark dataset, perform substrate interference suppression on the noise reduction coating thickness time series dataset to generate substrate interference-free coating time series data. S24. Perform multi-layer structure signal separation processing on the substrate interference removal coating timing data to generate pure coating thickness dynamic trace data.

6. The online X-ray real-time feedback control method for coating thickness in electroplating processes according to claim 1, characterized in that, The S3 includes: S31. Import the dynamic trace data of the pure coating thickness and generate the input dataset for the electrochemical deposition model calibration. S32. Based on the input dataset of the electrochemical deposition model, iteratively update the parameters of the electrochemical deposition model to generate the corrected electrochemical deposition model. S33. Calculate the local current density distribution data using the corrected electrochemical deposition model, and generate a local current density distribution dataset. S34. Calculate the coating growth rate field data based on the local current density distribution dataset, and generate the coating growth rate field dataset. S35. Perform nonlinear coupling analysis on the local current density distribution dataset and the coating growth rate field dataset to generate micro-area thickness deviation prediction data.

7. The online X-ray real-time feedback control method for coating thickness in electroplating processes according to claim 6, characterized in that, S33 includes: Extract the core parameters of the corrected electrochemical deposition model, and decompose the electrode reaction coefficient, electrolyte conductivity and interfacial impedance parameters to generate the core parameter set of the model. The core parameter set of the correlation model and the calibration input dataset of the electrochemical deposition model are used to extract the correlation features of micro-area coating growth and generate the input set for current density calculation. Based on the current density calculation input set, the point-by-point current density calculation of the micro-region is carried out through the corrected model to generate a preliminary local current density dataset. By comparing the correlation between the preliminary dataset and the dynamic trace data of coating thickness, abnormal calculation results are filtered out, and a current density correction dataset is generated. The spatial coordinates and numerical accuracy of the dataset are regularized and corrected to generate a local current density distribution dataset.

8. The online X-ray real-time feedback control method for coating thickness in electroplating processes according to claim 1, characterized in that, The S4 includes: S41. Import micro-area thickness deviation prediction data and workpiece motion trajectory data to generate spatial coordinate mapping analysis input dataset; S42. Perform full-dimensional coordinate matching on the input dataset based on spatial coordinate mapping analysis to generate a coordinate matching dataset; S43. Generate a thermal map of the thickness deviation of the entire workpiece surface using the coordinate matching dataset; S44. Perform edge effect and shading area feature identification on the thermal map of thickness deviation of the entire workpiece surface to generate local anomaly control area data; S45. Based on the data of the local abnormal control area, perform millisecond-level closed-loop correction of electroplating parameters to generate a real-time control instruction set.

9. The online X-ray real-time feedback control method for coating thickness in electroplating processes according to claim 1, characterized in that, The S5 includes: S51. Import the real-time control instruction set and generate a multi-actuator collaborative control input dataset; S52. Based on the multi-actuator collaborative control input dataset, drive the linkage operation of each actuator to generate a dynamic electroplating process parameter field; S53. Collect real-time monitoring data of coating thickness under dynamic electroplating process parameter field, and generate input dataset for stability index calculation; S54. Calculate the coating thickness stability index on the input dataset for stability index calculation, and generate comprehensive electroplating quality assessment data. S55. Based on comprehensive electroplating quality assessment data, a graded early warning mechanism is triggered to generate intelligent control feedback data for the electroplating process.

10. A system for implementing the online X-ray real-time feedback control method for coating thickness in the electroplating process as described in claim 1, characterized in that, The system includes: Network construction module: Divides the moving workpiece into micro-area coating thickness trace measurement regions, generating dynamic micro-area measurement grid data; deploys an X-ray fluorescence scanning device based on the dynamic micro-area measurement grid data to construct a high-resolution (≤1mm) network. 2 Online trace measurement network; Signal processing module: Based on the online trace measurement network, it performs millisecond-level synchronous data acquisition to generate a time-series dataset of coating thickness traces and workpiece motion trajectory data; it performs substrate interference suppression and multi-layer structure signal separation processing on the time-series dataset of coating thickness traces to generate clean dynamic trace data of coating thickness. Coupling analysis module: Based on the dynamic trace data of pure coating thickness, the electrochemical deposition model is calibrated online to obtain local current density distribution data and coating growth rate field data respectively; Nonlinear coupling analysis is performed on the coating growth rate field data through local current density distribution data to generate micro-area thickness deviation prediction data; Command control module: It uses workpiece motion trajectory data to perform spatial coordinate mapping processing on micro-area thickness deviation prediction data to generate a full workpiece surface thickness deviation heatmap; it uses the full workpiece surface thickness deviation heatmap to identify edge effects and shading areas, generating local abnormal control area data; it performs millisecond-level electroplating parameter closed-loop correction on the local abnormal control area data to generate a real-time control command set; The graded early warning module performs collaborative control of multiple actuators based on the real-time control instruction set to generate a dynamic electroplating process parameter field; it calculates the coating thickness stability index based on the dynamic electroplating process parameter field to generate comprehensive electroplating quality assessment data; and it triggers a graded early warning mechanism based on the comprehensive electroplating quality assessment data to generate intelligent control feedback data for the electroplating process.

Images