Automatic control method and device for FPC production

Abstract

The invention relates to the technical field of flexible circuit board production, and discloses an automatic control method and device for FPC production, and the method comprises the steps: collecting the material characteristics and environmental condition data of production equipment through a sensor, and carrying out the preprocessing, and obtaining an analysis data flow; carrying out feature value extraction and standardization processing on the data to form a standardized feature vector set, evaluating parameter deviation of the production equipment according to the standardized feature vector set, and obtaining a preliminary evaluation result; and when the result exceeds a preset threshold value, quantitatively analyzing the line precision and the interlayer structure by using a deviation analysis algorithm to obtain specific distribution data causing deviation. And performing parameter matching calculation in combination with the current production state, designing a dynamic adjustment scheme, and updating processing time and energy setting in real time to obtain first production configuration data so as to complete overall parameter configuration. According to the method, production monitoring is enhanced, adaptive adjustment of parameters is realized, the consistency of FPC production quality is effectively improved, and reliable technical support is provided for flexible electronic manufacturing.

Description

Technical Field

[0001] The invention relates to the technical field of flexible circuit board production, and discloses an automatic control method and device for FPC production. Background Art

[0002] Amid the rapid growth of the electronic information industry, flexible printed circuits (FPCs), with their lightweight, flexible nature and flexibility, are widely used in smartphones, wearable devices, aerospace, and other fields, becoming a core component of modern electronic manufacturing. Their production quality not only directly determines the performance and reliability of end products but also serves as a key constraint on technological innovation and upgrading in the electronics industry. As market demand for smaller and more powerful electronic products continues to rise, the scale of FPC production continues to expand, and the requirements for precise control of the production process are becoming increasingly stringent, creating an urgent need for advanced production control technologies to support industry development.

[0003] Currently, most flexible circuit board production process control technologies rely on traditional models, pre-set production parameters and fixed process flows. While some technologies incorporate sensors for data collection and combine simple algorithms to analyze and process production data, they still have significant shortcomings when faced with complex and volatile production environments and fluctuating material properties. These methods often focus on post-production testing and correction, performing quality inspections on already manufactured products and then adjusting production parameters based on the test results. However, they lack the ability to anticipate and proactively intervene in potential production process issues.

[0004] In the actual production of flexible circuit boards, differences in material properties, changes in ambient temperature and humidity, and fluctuations in equipment operation interact with each other, causing production parameters such as processing time and energy settings to easily deviate from the optimal values. Parameter deviations can cause problems such as line accuracy errors and interlayer structural misalignment, resulting in unstable product quality and increased scrap rates. Existing technologies are difficult to accurately control the production process in real time and cannot adapt to complex scenarios with multi-parameter coupling, resulting in low production efficiency and poor product consistency, which seriously hinders the high-quality development of the flexible circuit board industry. In summary, how to monitor key parameters in real time during the production process and actively compensate for deviations through dynamic adjustments has become a key issue in improving the production quality and consistency of flexible circuit boards. Summary of the Invention

[0005] The present invention provides an automatic control method and device for FPC production, so as to enhance production monitoring, realize adaptive parameter adjustment, and provide reliable technical support for flexible electronic manufacturing.

[0006] In a first aspect, in order to solve the above technical problems, the present invention provides an automatic control method for FPC production, comprising: Collect material properties and environmental condition data in production equipment through sensors, and pre-process the collected data to obtain analytical data streams; Extracting eigenvalues from the analysis data stream, and performing standardization processing on the extracted eigenvalues to form a standardized eigenvector set; Evaluate the parameter deviation of the production equipment according to the feature vector set to obtain a preliminary evaluation result; When the preliminary evaluation result exceeds the preset evaluation threshold standard, a preset deviation analysis algorithm is used to quantitatively analyze the line accuracy and interlayer structure to obtain specific distribution data that causes the deviation; Perform parameter matching calculations and design dynamic adjustment plans based on the specific distribution data and current production status; According to the dynamic adjustment scheme, the processing time and energy settings are updated in real time to obtain the adjusted first production configuration data; Complete overall parameter configuration according to the first production configuration data.

[0007] In an optional embodiment, extracting feature values from the analysis data stream and performing standardization on the extracted feature values to form a standardized feature vector set includes: Segmenting the analysis data stream and calculating statistical characteristic values of each segment of data; wherein the statistical characteristic values include but are not limited to mean, standard deviation, maximum value, and minimum value; applying a frequency domain transformation algorithm to the analysis data stream according to the statistical characteristic values to extract periodic characteristic parameters; Based on the periodic characteristic parameters and historical production data, the characteristic values are normalized using the Z-score standardization method; The normalized eigenvalues are dimensionally transformed to generate a set of eigenvectors containing multidimensional features of material properties and environmental conditions.

[0008] In an optional embodiment, evaluating the parameter deviation of the production equipment according to the feature vector set to obtain a preliminary evaluation result includes: Based on the feature vector set, a parameter deviation evaluation model is constructed, and the weight coefficient of each dimension feature is set; According to the parameter deviation evaluation model, the Euclidean distance between the current feature vector and the standard production state feature vector is calculated and used as a preliminary deviation indicator; By using a preset proofreading module, the preliminary deviation indicator is compared with the historical deviation data to generate the preliminary evaluation result; The preliminary assessment results include the degree of deviation, the type of deviation and the trend of deviation.

[0009] In an optional embodiment, the quantitative analysis of the line accuracy and interlayer structure using a preset deviation analysis algorithm to obtain specific distribution data causing the deviation includes: Through the deviation analysis algorithm, the line accuracy error and inter-layer structure misalignment are quantified to generate basic deviation distribution data; Based on the deviation distribution basic data, a regression analysis algorithm is used to decompose the line accuracy error data and determine the main influencing factors; Through image recognition technology, three-dimensional modeling analysis is performed on the structural dislocation data to determine the dislocation location; The dislocation positions are aligned with the main influencing factors, and basic corrections are performed through weighted fusion to obtain the specific distribution data.

[0010] In an optional embodiment, performing parameter matching calculation and designing a dynamic adjustment plan based on the specific distribution data and the current production status includes: Based on the specific distribution data, the correlation between the concentrated processing time and energy settings of the abnormal deviation points is analyzed, the processing time is prioritized according to the degree of deviation, and a time control adjustment list is formed; Combined with the current production status data, parameter matching basis is extracted from the time control adjustment list, and the parameters are classified by a support vector machine model to determine the optimal parameter combination; Extracting energy setting optimization requirements from the optimal parameter combination, generating a preliminary energy adjustment plan, and obtaining an energy optimization configuration; Analyze the dynamic adjustment execution conditions of the energy optimization configuration, formulate specific steps for dynamic adjustment, and determine the adjustment execution sequence; According to the adjustment execution sequence, the dynamic adjustment plan is obtained in combination with the control requirements and the execution plan logic mapping.

[0011] In an optional embodiment, updating the processing time and energy settings in real time according to the dynamic adjustment scheme to obtain the adjusted first production configuration data includes: According to the dynamic adjustment scheme, real-time operation data is obtained from the equipment, processing time and energy settings are analyzed, and preliminary deviation assessment results are obtained; Based on the preliminary deviation assessment results, the processing time and energy settings are optimized and analyzed through a preset decision model to determine the adjustment direction and amplitude and generate preliminary control instructions; Sending the preliminary control instructions to the target device, updating the device operation status in real time and obtaining operation feedback data; If there is a deviation between the operation feedback data and the expected optimization target, a regression analysis algorithm is used to perform a secondary calibration on the processing to obtain a calibrated control instruction; Adjust production parameter configuration according to calibrated control instructions, continuously monitor equipment operating status, and obtain adjusted parameter configuration data; The first production configuration data is determined according to the parameter configuration data and the optimization target of the dynamic adjustment scheme.

[0012] In an optional embodiment, after completing the overall parameter configuration according to the first production configuration data, the method further includes: Analyze the real-time monitoring capabilities of the equipment through a preset loop feedback mechanism to determine the updated monitoring accuracy level; According to the monitoring accuracy level, the overall parameters are corrected twice using a preset deviation compensation algorithm to obtain second production configuration data; Complete the overall parameter configuration according to the second production configuration data.

[0013] In an optional embodiment, performing secondary correction on the overall parameters using a preset deviation compensation algorithm according to the monitoring accuracy level to obtain the second production configuration data includes: According to the monitoring accuracy level, the production parameter data is collected in real time, and pre-processed using a mean filter algorithm to obtain a smoothed parameter data set; Performing trend fitting on the parameter data set and the preset deviation standard by a linear regression algorithm to construct a deviation prediction model; Calculating compensation values of residual parameters according to the deviation prediction model to form a preliminary correction parameter set; Performing a consistency check on the preliminary calibration parameter set to obtain test data; Based on the test data, the preliminary correction parameter set is secondary corrected using a compensation value combined with a gradient descent algorithm to optimize the residual deviation to obtain an optimized parameter set; The optimized parameter set is determined as second production configuration data.

[0014] In a second aspect, the present invention provides an automatic control device for FPC production, comprising: The data acquisition module is used to collect material characteristics and environmental conditions data in the production equipment through sensors, and pre-process the collected data to obtain analysis data streams; A standardization processing module, configured to extract feature values from the analysis data stream and perform standardization processing on the extracted feature values to form a standardized feature vector set; a deviation evaluation module, configured to evaluate the parameter deviation of the production equipment based on the feature vector set to obtain a preliminary evaluation result; a deviation analysis module configured to, when the preliminary evaluation result exceeds a preset evaluation threshold, quantitatively analyze the line accuracy and interlayer structure using a preset deviation analysis algorithm to obtain specific distribution data causing the deviation; A parameter matching module is used to perform parameter matching calculations and design dynamic adjustment plans based on the specific distribution data and the current production status; A first adjustment module is configured to update the processing time and energy settings in real time according to the dynamic adjustment scheme to obtain adjusted first production configuration data; The first configuration module is used to complete overall parameter configuration according to the first production configuration data.

[0015] In an optional embodiment, the device further includes: A feedback adjustment module is used to analyze the real-time monitoring capabilities of the equipment through a preset loop feedback mechanism to determine the updated monitoring accuracy level; a second adjustment module, configured to perform a secondary correction on the overall parameters according to the monitoring accuracy level using a preset deviation compensation algorithm to obtain second production configuration data; The second configuration module is used to complete the overall parameter configuration according to the second production configuration data.

[0016] Compared with the prior art, the present invention has the following beneficial effects: (1) Extract eigenvalues from the analysis data stream and perform standardization on the extracted eigenvalues to form a standardized feature vector set. This step uses algorithms such as sliding windows and fast Fourier transforms to accurately extract time domain and frequency domain features from the analysis data stream, and then eliminates dimensional differences through standardization methods such as Z scores. This ensures that the feature vector set can comprehensively and uniformly reflect the characteristics of production data, providing a standardized and accurate data foundation for subsequent analysis, and effectively improving the reliability and stability of data analysis.

[0017] (2) Based on the feature vector set, the parameter deviation of the production equipment is evaluated to obtain preliminary evaluation results. Using principal component analysis, clustering algorithms and other techniques, the feature vector set is compared and analyzed with the standard parameter characteristics to quantify the degree of deviation of the equipment parameters from the ideal state. Potential parameter anomalies can be quickly identified to provide a basis for production anomaly warning, avoid product quality degradation due to accumulated parameter deviations, and improve the controllability of the production process.

[0018] (3) Quantitatively analyze the line accuracy and interlayer structure using a preset deviation analysis algorithm to obtain specific distribution data that causes the deviation. Using regression analysis and machine learning feature extraction algorithms, combined with production process parameters and equipment operation data, we can deeply analyze the causes of line accuracy errors and interlayer structure misalignment. Accurately locate the location of deviations and influencing factors, facilitating targeted adjustments to the production process, reducing scrap rates, and ensuring that key product indicators meet standards.

[0019] (4) Based on the specific distribution data and the current production status, parameter matching calculations are performed and a dynamic adjustment plan is designed. Taking into account multiple factors such as deviation distribution, equipment load, and order demand, an intelligent optimization algorithm is used to establish a parameter matching model. Combined with actual production constraints, a dynamic adjustment plan adapted to the current working conditions is generated, enabling production parameters to quickly respond to production changes and improving the production system's adaptability and production efficiency.

[0020] (5) According to the dynamic adjustment plan, the processing time and energy settings are updated in real time to obtain the adjusted first production configuration data. Through the equipment control interface, the dynamic adjustment plan is converted into an executable instruction for the equipment, and key parameters such as processing time and energy are modified in real time. This achieves instant optimization of production parameters, reduces the lag in parameter adjustment, ensures that the production process is continuously efficient and stable, and improves product production quality and consistency.

[0021] (6) Based on the monitoring accuracy level, the overall parameters are secondary corrected using a preset deviation compensation algorithm to obtain the second production configuration data. Based on the accuracy indicators of real-time monitoring, algorithms such as gradient descent and deviation compensation models are used to deeply optimize the overall production parameters. Minor deviations caused by factors such as environmental fluctuations and equipment wear are corrected to further improve the accuracy of production parameters, ensure that product quality reaches higher standards, and enhance the stability and reliability of the production process. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 1 is a flow chart of an automatic control method for FPC production provided by an embodiment of the present invention; Figure 2 This is a schematic structural diagram of an automatic control device for FPC production provided by an embodiment of the present invention; Figure 3 This is another structural schematic diagram of an automatic control device for FPC production provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0023] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0024] First, refer to Figure 1 , an embodiment of the present invention provides an automatic control method for FPC production, comprising the following steps: S11, collecting material properties and environmental condition data in the production equipment through sensors, and preprocessing the collected data to obtain an analysis data stream; S12, extracting feature values from the analysis data stream, and performing standardization processing on the extracted feature values to form a standardized feature vector set; S13, evaluating parameter deviations of production equipment based on the feature vector set to obtain preliminary evaluation results; S14, when the preliminary evaluation result exceeds a preset evaluation threshold, quantitatively analyzing the line accuracy and interlayer structure using a preset deviation analysis algorithm to obtain specific distribution data causing the deviation; S15, performing parameter matching calculations and designing a dynamic adjustment plan based on the specific distribution data and the current production status; S16, updating the processing time and energy settings in real time according to the dynamic adjustment scheme to obtain adjusted first production configuration data; S17: Complete overall parameter configuration according to the first production configuration data.

[0025] In step S11, material properties and environmental condition data in the production equipment are collected by sensors, and the collected data are preprocessed to obtain an analysis data stream.

[0026] In one implementation, using the flexible circuit board (FPCB) manufacturing process as an example, a sensor network is deployed to collect and preprocess production data. Ten laser displacement sensors with a resolution of 0.01mm are installed on the placement machine, measuring substrate thickness in real time at a 100Hz acquisition frequency, generating 1,000 data points per second. A DHT22 temperature and humidity sensor is embedded in the environmental control system, recording workshop temperature (20-25°C) and humidity (40-60% RH) at 1-second intervals, with a temperature accuracy of ±0.5°C and a humidity accuracy of ±2% RH. The collected data is processed by edge computing nodes, with data packets under 1KB for low-latency transmission. The data is then distributed to an analysis platform via Apache Kafka. A standard deviation threshold of 0.05mm is set for substrate thickness data, triggering an anomaly alert when 10 consecutive measurements exceed the threshold. A K-means clustering algorithm (k=3) is used, with k-means++ initializing cluster centers and 100 iterations, to classify anomaly data to identify sources such as placement deviation, material defects, or environmental interference. If the chip offset rate is detected to exceed 0.1mm, the placement machine speed is automatically reduced by 10%. Finally, the processed data is integrated into an analytical data stream for subsequent production parameter analysis and control. This implementation achieves efficient collection and precise processing of production data, significantly improving production stability and product qualification rate.

[0027] In step S12, feature value extraction is performed on the analysis data stream, and the extracted feature values are standardized to form a standardized feature vector set.

[0028] In one embodiment, the analysis data stream is segmented and the statistical characteristic values of each segment of data are calculated; the statistical characteristic values include but are not limited to the mean, standard deviation, maximum value, and minimum value; based on the statistical characteristic values, a frequency domain transformation algorithm is applied to the analysis data stream to extract periodic characteristic parameters; based on the periodic characteristic parameters and historical production data, the characteristic values are normalized using the Z-score standardization method; the normalized characteristic values are dimensionalized to generate a characteristic vector set containing multidimensional characteristics of material properties and environmental conditions.

[0029] Specifically, using the flexible circuit board production process as an example, the analysis data stream is processed to generate a set of feature vectors. Assume that the real-time data stream contains material property data including tensile strength (unit: MPa) and environmental condition data including temperature (unit: °C). The data is collected once per second, resulting in a tensile strength sequence of [450.5, 451.2, 449.8, 452.1] and a temperature sequence of [25.3, 25.7, 26.1, 25.9]. First, the analysis data stream is segmented using a sliding window technique with a window size of 4 and a step size of 1. The statistical eigenvalues of the segmented tensile strength sequence are calculated, resulting in a mean of 450.65 MPa and a standard deviation of 1.03 MPa. The maximum value of the sequence is 452.1 MPa, and the minimum value is 449.8 MPa. Next, based on the aforementioned statistical eigenvalues, a Fast Fourier Transform (FFT) frequency domain algorithm was applied to the temperature series to extract periodic characteristic parameters. The primary frequency was identified as 0.25 Hz, indicating a temperature fluctuation period of approximately 4 seconds. The eigenvalues were then normalized using the Z-score method, combined with historical production data. Given the historical mean of tensile strength of 450 MPa and standard deviation of 2 MPa, the Z-score was calculated to be (450.65 - 450) / 2 = 0.325. The historical mean of temperature frequency was 0.2 Hz and standard deviation was 0.33 Hz, resulting in a normalized Z-score of 0.15. Finally, the normalized eigenvalues, including the mean, standard deviation, maximum, minimum values of tensile strength, and temperature frequency, were transformed into a dimensionality-based set. The multidimensional features of material properties (tensile strength-related features) and environmental conditions (temperature frequency features) were integrated to generate a standardized set of feature vectors for subsequent production equipment parameter deviation assessment. This embodiment improves the accuracy of feature vectors through standardization and multidimensional integration, ensuring the reliability of subsequent analysis.

[0030] In step S13, the parameter deviation of the production equipment is evaluated based on the feature vector set to obtain a preliminary evaluation result.

[0031] In one embodiment, a parameter deviation evaluation model is constructed based on the feature vector set, and the weight coefficient of each dimensional feature is set; based on the parameter deviation evaluation model, the Euclidean distance between the current feature vector and the standard production state feature vector is calculated and used as a preliminary deviation indicator; through a preset proofreading module, the preliminary deviation indicator is compared with the historical deviation data to generate the preliminary evaluation result; wherein, the preliminary evaluation result includes the degree of deviation, deviation type and deviation trend.

[0032] Specifically, taking the flexible circuit board production process as an example, based on the generated feature vector set [0.325, 0.515, 1.05, -0.1, 0.15] (including the tensile strength mean, standard deviation, maximum value, minimum value, and temperature frequency Z score), the specific steps for parameter deviation evaluation are as follows: Construct a parameter deviation evaluation model and set the weight coefficients of each dimension feature: tensile strength mean 0.3, standard deviation 0.2, maximum 0.1, minimum 0.1, and temperature frequency 0.3. Based on this model, calculate the weighted Euclidean distance between the current feature vector and the standard production state feature vector (assuming it is [0, 0, 0, 0, 0]): , where n = 5 is the feature dimension, w i is the weight coefficient of the i-th dimension feature, x i is the Z score of the current eigenvalue, y i is the Z score of the characteristic value of the standard production state (usually 0). Substitute the specific numerical value to calculate: ; 0.42 is used as the preliminary deviation indicator.

[0033] Compare the preliminary deviation indicators with the historical deviation data through the preset proofreading module: Deviation degree: The threshold is set at 0.5, and 0.42 < 0.5 is considered as mild deviation; Deviation type: By analyzing the contribution of each dimension (e.g., the standard deviation of tensile strength has the highest contribution), it is determined that the deviation is caused by fluctuations in material properties. Deviation trend: Comparing the results of the last 10 evaluations (assuming the historical deviation values decrease in sequence: 0.55 → 0.48 → 0.42), it is determined to be a convergence trend.

[0034] The resulting preliminary assessment results showed slight material property fluctuations (deviation value 0.42) and a convergent trend. This result can be used to guide subsequent parameter adjustments, such as maintaining the current heating power fine-tuning strategy and continuously monitoring changes in tensile strength.

[0035] In step S14, when the preliminary evaluation result exceeds a preset evaluation threshold, a preset deviation analysis algorithm is used to perform a quantitative analysis on the line accuracy and interlayer structure to obtain specific distribution data causing the deviation.

[0036] In one embodiment, a deviation analysis algorithm is used to quantify line accuracy errors and interlayer structural misalignment to generate basic deviation distribution data; based on the deviation distribution basic data, a regression analysis algorithm is used to decompose line accuracy error data to determine the main influencing factors; through image recognition technology, three-dimensional modeling analysis is performed on the structural misalignment data to determine the misalignment position; the misalignment position is aligned with the main influencing factors, and basic corrections are performed through weighted fusion to obtain the specific distribution data.

[0037] Specifically, taking the quantitative analysis of line accuracy error and interlayer structure misalignment in the production process of flexible circuit boards as an example, the specific implementation steps are as follows: First, a deviation analysis algorithm was activated to quantitatively analyze key quality parameters. A high-precision sensor network was deployed to collect real-time data. Laser displacement sensors (resolution 0.001mm) collected actual line coordinate data and simultaneously recorded processing equipment operating parameters (e.g., current average processing time of 12.5 seconds, standard deviation 0.8 seconds, target processing time of 12.0 seconds, deviation 0.5 seconds). An industrial camera (30fps) was used to capture interlayer structural images, and an infrared positioning module was used to obtain 3D coordinate information. A normal distribution analysis of the processing time data revealed that 68% of processing times fell within the range of 11.7 seconds to 13.3 seconds, with a 32% probability of exceeding the target range (11.5 seconds to 12.5 seconds). This provided a foundation for deviation assessment.

[0038] It is worth noting that the deviation analysis algorithm includes a multi-dimensional quantitative model and a data fusion module: Multi-dimensional quantification model: To address line accuracy errors, a least-squares linear fitting algorithm was used to compare the actual line coordinates (e.g., the actual X-axis coordinates of a batch of samples [1.002mm, 1.005mm, 0.998mm]) with the designed coordinates [1.000mm, 1.000mm, 1.000mm] point by point. The mean absolute value of the calculated error was 0.003mm, with a standard deviation of 0.002mm. To address interlayer structural misalignment, a three-dimensional coordinate transformation algorithm (based on the spatial rectangular coordinate system transformation matrix) was used to unify the coordinates of feature points on different layers (e.g., the top feature point (5.00mm, 3.00mm, 0.1mm) and the bottom feature point (5.02mm, 3.01mm, 0.1mm)) to the reference coordinate system. The calculated mean X-direction misalignment was 0.02mm, the Y-direction misalignment was 0.015mm, and the Z-direction interlayer thickness deviation was 0.005mm.

[0039] The data fusion module performs Z-score normalization on processing time deviation data (average deviation 0.5 seconds, standard deviation 0.8 seconds), line accuracy error data (mean 0.003mm, standard deviation 0.002mm), and interlayer structure misalignment data (mean X / Y misalignment). (Based on three months of historical production data, the calculated processing time mean μ = 12.1 seconds, σ = 0.7 seconds; line accuracy error μ = 0.002mm, σ = 0.001mm) yields normalized values of 0.57, 1.0, and 2.0, respectively. Using a weighted summation algorithm (processing time weight 0.3, line accuracy weight 0.4, and interlayer misalignment weight 0.3), the comprehensive deviation index is calculated as 0.3 × 0.57 + 0.4 × 1.0 + 0.3 × 2.0 = 0.97, generating basic deviation distribution data.

[0040] Subsequently, an in-depth analysis was conducted based on the basic data of deviation distribution: a multivariate linear regression algorithm (R²=0.92) was used to model the line accuracy error data. The input variables were processing time, energy setting, and ambient temperature, and the output was the line accuracy error value. A significance test (P<0.05) determined that energy setting (coefficient 0.62) and processing time (coefficient 0.38) were the main influencing factors. At the same time, image recognition technology was used to perform three-dimensional modeling of the interlayer structure misalignment data: 200 multi-angle images of the edge area of the circuit board were collected through a binocular vision system, and a three-dimensional model was generated through SIFT feature matching and point cloud reconstruction. The concentrated area of misalignment was located at the edge zone 0-2mm from the edge of the board, accounting for 78% of the total misalignment area.

[0041] Finally, the misalignment locations determined by 3D modeling were spatially and parametrically aligned with the main influencing factors identified by regression analysis: The energy setting weight was set at 0.4, and the processing time weight was set at 0.6. A weighted fusion algorithm was then used to perform basic corrections to the deviation data (for example, when the energy setting was adjusted from 1500 joules to 1450 joules, the mean line accuracy error dropped to 0.002mm; when the processing time was shortened to 12.2 seconds, the amount of misalignment in the edge area was reduced by 30%). This comprehensive data was generated, including specific error values (line accuracy error 0.003mm, inter-layer misalignment 0.02mm), influencing factor weights (energy 0.4, time 0.6), and misalignment locations (0-2mm area at the edge). This provided precise data support for subsequent parameter adjustments, improving production quality consistency to over 92%.

[0042] In step S15, parameter matching calculation is performed based on the specific distribution data and the current production status, and a dynamic adjustment plan is designed; In one embodiment, based on the specific distribution data, the correlation between the concentrated processing time and energy setting of the abnormal deviation points is analyzed, the processing time is prioritized according to the degree of deviation, and a time control adjustment list is formed; in combination with the current production status data, the parameter matching basis is extracted from the time control adjustment list, and the parameters are classified through a support vector machine model to determine the optimal parameter combination; the energy setting optimization requirements are extracted from the optimal parameter combination, and a preliminary energy adjustment plan is generated to obtain an energy optimization configuration; the dynamic adjustment execution conditions of the energy optimization configuration are analyzed, specific steps for dynamic adjustment are formulated, and an adjustment execution sequence is determined; according to the adjustment execution sequence, the control requirements and the execution plan logical mapping are combined to obtain the dynamic adjustment plan.

[0043] Specifically, using flexible circuit board production as an example, the system uses sensors to collect real-time processing equipment operating parameters and obtain processing time deviation data. The system reports an average processing time of 12.5 seconds, a standard deviation of 0.8 seconds, and a calculated deviation of 0.5 seconds. The target value is 12.0 seconds, and 68% of processing times fall between 11.7 and 13.3 seconds, with a probability of exceeding the target range of 32%. Energy consumption data is also collected. The current average energy setting is 1500 joules. Historical data regression analysis shows that an energy setting of 1450 joules improves processing quality scores by 5% and reduces energy consumption by 3.2%. After generating specific distribution data, the system analyzes the correlation coefficient between processing time and energy setting at the outlier deviation points and finds a 0.78 correlation. Processing time is ranked by deviation level, with shortening processing time being the top optimization priority. This results in a time control adjustment list. By combining production status data such as current equipment load and order urgency, the system extracts parameter matching criteria from the list. Using a support vector machine model, the system classifies parameters such as processing time and energy setting, determining the optimal parameter combination: a processing time of 12.2 seconds and an energy setting of 1460 joules. Energy optimization requirements are then extracted from this combination, a preliminary energy adjustment plan is generated, and an energy-optimized configuration is obtained. Subsequently, the system sets a trigger for adjustment when the processing time deviation exceeds 0.3 seconds or the energy deviation exceeds 50 joules. This system establishes an execution sequence that first adjusts the processing time parameters and then simultaneously updates the energy settings. Finally, based on this sequence, the control requirements are logically mapped to the execution plan to obtain the final execution plan data, including the parameter adjustment target values, execution sequence, and trigger conditions, thus achieving dynamic optimization and closed-loop control of production parameters.

[0044] In step S16, according to the dynamic adjustment scheme, the processing time and energy settings are updated in real time to obtain the adjusted first production configuration data; In one embodiment, according to the dynamic adjustment scheme, real-time operation data is obtained from the equipment, the processing time and energy settings are analyzed, and a preliminary deviation evaluation result is obtained; according to the preliminary deviation evaluation result, the processing time and energy settings are optimized and analyzed through a preset decision model, the adjustment direction and amplitude are determined, and a preliminary control instruction is generated; the preliminary control instruction is sent to the target equipment, the equipment operation status is updated in real time and operation feedback data is obtained; if there is a deviation between the operation feedback data and the expected optimization target, the regression analysis algorithm is used to perform a secondary calibration on the processing to obtain a calibrated control instruction; according to the calibrated control instruction, the production parameter configuration is adjusted, the equipment operation status is continuously monitored, and the adjusted parameter configuration data is obtained; according to the parameter configuration data and the optimization target of the dynamic adjustment scheme, the first production configuration data is determined.

[0045] Specifically, a dynamic adjustment mechanism was implemented in the flexible circuit board production scenario. A real-time data acquisition system collected real-time operating data from equipment sensors, including a processing speed of 50 pieces / minute, an operating power of 2000 watts, and an ambient temperature of 25°C. Analysis revealed that the current processing speed deviated by 10 pieces / minute from the ideal value of 60 pieces / minute, and that energy utilization was only 75% at 2000 watts. This yielded a preliminary deviation assessment. Based on this, the system invoked a pre-set decision model for optimization analysis, determining to increase the equipment speed to 1100 rpm to adjust the processing time while simultaneously reducing the power to 1800 watts to optimize energy settings. Preliminary control instructions were generated and issued to the equipment. After the equipment executed the instructions, the system collected operational feedback data. If the feedback indicated that the processing speed had only increased to 58 pieces / minute and the energy utilization was 82%, falling short of the expected targets, a regression analysis algorithm was used for secondary calibration, fine-tuning the speed to 1120 rpm and the power to 1780 watts, resulting in calibrated control instructions. The production parameter configuration is updated accordingly, and the equipment operating status is continuously monitored. Finally, combined with the efficiency and energy consumption optimization goals of the dynamic adjustment plan, the first production configuration data with a processing speed of 60 pieces / minute, a rotation speed of 1120 rpm, and a power of 1780 watts is determined, achieving closed-loop optimization of production parameters.

[0046] In a second aspect, an embodiment of the present invention further provides an automatic control method for FPC production, which, after completing the overall parameter configuration according to the first production configuration data, further includes: Analyze the real-time monitoring capabilities of the equipment through a preset loop feedback mechanism to determine the updated monitoring accuracy level; According to the monitoring accuracy level, the overall parameters are corrected twice using a preset deviation compensation algorithm to obtain second production configuration data; Complete the overall parameter configuration according to the second production configuration data.

[0047] In one embodiment, the overall parameters are secondary corrected according to the monitoring accuracy level through a preset deviation compensation algorithm to obtain second production configuration data, including: according to the monitoring accuracy level, production parameter data is collected in real time, and preprocessed using a mean filtering algorithm to obtain a smoothed parameter data set; trend fitting is performed on the parameter data set and a preset deviation standard through a linear regression algorithm to construct a deviation prediction model; compensation values of residual parameters are calculated according to the deviation prediction model to form a preliminary correction parameter set; consistency test is performed on the preliminary correction parameter set to obtain test data; based on the test data, the preliminary correction parameter set is secondary corrected using the compensation value combined with the gradient descent algorithm to optimize the residual deviation to obtain an optimized parameter set; and the optimized parameter set is determined as the second production configuration data.

[0048] Specifically, the system analyzes the real-time monitoring capabilities of the equipment through a preset loop feedback mechanism: it uses the PID control algorithm (P=0.5, I=0.2, D=0.1) to dynamically adjust production parameters, reducing the temperature deviation from ±0.5°C to ±0.2°C, and optimizing the pressure deviation from ±0.3MPa to ±0.1MPa; at the same time, it adaptively adjusts the acquisition frequency (maintained at 1 time per minute) based on the data volatility (the current temperature volatility is 0.014, and the pressure volatility is 0.02). The error analysis calculates that the root mean square error (RMSE) is reduced from 0.6 to 0.3, and it is determined that the updated monitoring accuracy level has increased by 50%, meeting the preset standard of error less than 0.5%. Based on this level of monitoring accuracy, a secondary correction is performed using a preset deviation compensation algorithm: real-time data collection of production parameters such as temperature (28.0°C) and pressure (2.5MPa) is performed, preprocessed using a mean filter algorithm, and a smoothed parameter data set is obtained. A linear regression algorithm is used to perform trend fitting on the data set and the preset deviation standards (temperature deviation threshold ±0.1°C, pressure deviation threshold ±0.05MPa). A deviation prediction model is constructed, and compensation values of 0.3°C for temperature residual deviation and 0.2MPa for pressure residual deviation are calculated, forming a preliminary correction parameter set (temperature target 28.2°C, pressure target 2.45MPa). The preliminary correction parameter set is subjected to a consistency test to verify its match with the production quality target (for example, the line accuracy error must be ≤±0.03mm). After obtaining the test data, a secondary correction is performed using the compensation value combined with the gradient descent algorithm. The gradient descent iterative formula is supplemented when optimizing the residual deviation: , where θt is the correction parameter for the tth iteration (e.g. the initial value of the temperature correction coefficient is set to 1.02, the initial value of the pressure correction coefficient is set to 0.98), α is the learning rate (e.g. 0.01), is the gradient of the loss function (the loss function is defined as the sum of squared residual deviations, i.e. (J = (actual temperature − target temperature) 2 + (actual pressure − target pressure) 2), where the gradient is the partial derivative of the loss function with respect to the correction parameter).

[0049] When performing secondary optimization on the preliminary calibration parameter set, taking the temperature parameter as an example, in the first iteration: Initial temperature correction factor , the corresponding corrected temperature is 28.0°C × 1.02 = 28.56°C (exceeding the target by 28.2°C, with a residual deviation of 0.36°C); Calculate the loss function gradient = 2 × (corrected temperature − target temperature) × actual temperature = 2 × 0.36 × 28.0 = 20.16; Iterate through the gradient descent formula: The corrected temperature is 28.0°C × 0.8184 ≈ 22.92°C (the deviation is too large and further iteration is required).

[0050] After five iterations, the temperature correction coefficient converged to 1.0036, resulting in a corrected temperature of 28.0°C × 1.0036 ≈ 28.1°C (with a residual deviation of 0.1°C). Similarly, the pressure correction coefficient converged to 0.992 after three iterations, resulting in a corrected pressure of 2.5 MPa × 0.992 ≈ 2.48 MPa (with a residual deviation of 0.02 MPa). This ultimately resulted in the optimized parameter set, which was determined as the second production configuration data. Based on this second production configuration data, overall parameter configuration was completed, stabilizing the temperature within ±0.1°C and the pressure within ±0.05 MPa. Production quality consistency increased from 85% to 92%, ensuring high-precision control of the production process.

[0051] refer to Figure 2 The second embodiment of the invention provides an automatic control device for FPC production, comprising: The data acquisition module is used to collect material characteristics and environmental conditions data in the production equipment through sensors, and pre-process the collected data to obtain analysis data streams; A standardization processing module, configured to extract feature values from the analysis data stream and perform standardization processing on the extracted feature values to form a standardized feature vector set; a deviation evaluation module, configured to evaluate the parameter deviation of the production equipment based on the feature vector set to obtain a preliminary evaluation result; a deviation analysis module configured to, when the preliminary evaluation result exceeds a preset evaluation threshold, quantitatively analyze the line accuracy and interlayer structure using a preset deviation analysis algorithm to obtain specific distribution data causing the deviation; A parameter matching module is used to perform parameter matching calculations and design dynamic adjustment plans based on the specific distribution data and the current production status; A first adjustment module is configured to update the processing time and energy settings in real time according to the dynamic adjustment scheme to obtain adjusted first production configuration data; The first configuration module is used to complete overall parameter configuration according to the first production configuration data.

[0052] Reference Figure 3 In one embodiment, the device further comprises: A feedback adjustment module is used to analyze the real-time monitoring capabilities of the equipment through a preset loop feedback mechanism to determine the updated monitoring accuracy level; a second adjustment module, configured to perform a secondary correction on the overall parameters according to the monitoring accuracy level using a preset deviation compensation algorithm to obtain second production configuration data; The second configuration module is used to complete the overall parameter configuration according to the second production configuration data.

[0053] It should be noted that the automatic control device for FPC production provided in an embodiment of the present invention is used to execute all the process steps of the automatic control method for FPC production in the above embodiment. The working principles and beneficial effects of the two correspond one to one, so they will not be repeated here.

[0054] It should be noted that the device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed across multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment. In addition, in the drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines. A person of ordinary skill in the art can understand and implement the present invention without inventive effort.

[0055] The specific embodiments described above further illustrate the objectives, technical solutions, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. An automatic control method for FPC production, characterized in that, include: Collect material properties and environmental condition data in production equipment through sensors, and pre-process the collected data to obtain analytical data streams; Extracting eigenvalues from the analysis data stream, and performing standardization processing on the extracted eigenvalues to form a standardized eigenvector set; Evaluate the parameter deviation of the production equipment according to the feature vector set to obtain a preliminary evaluation result; When the preliminary evaluation result exceeds the preset evaluation threshold standard, a preset deviation analysis algorithm is used to quantitatively analyze the line accuracy and interlayer structure to obtain specific distribution data that causes the deviation; Perform parameter matching calculations and design dynamic adjustment plans based on the specific distribution data and current production status; According to the dynamic adjustment scheme, the processing time and energy settings are updated in real time to obtain the adjusted first production configuration data; Complete overall parameter configuration according to the first production configuration data.

2. The automatic control method for FPC production according to claim 1, characterized in that: The extracting characteristic values from the analysis data stream and performing standardization processing on the extracted characteristic values to form a standardized characteristic vector set includes: Segmenting the analysis data stream and calculating statistical characteristic values of each segment of data; wherein the statistical characteristic values include but are not limited to mean, standard deviation, maximum value, and minimum value; applying a frequency domain transformation algorithm to the analysis data stream according to the statistical characteristic values to extract periodic characteristic parameters; Based on the periodic characteristic parameters and historical production data, the characteristic values are normalized using the Z-score standardization method; The normalized eigenvalues are dimensionally transformed to generate a set of eigenvectors containing multidimensional features of material properties and environmental conditions.

3. The automatic control method for FPC production according to claim 1, characterized in that: The parameter deviation of the production equipment is evaluated based on the feature vector set to obtain a preliminary evaluation result, including: Based on the feature vector set, a parameter deviation evaluation model is constructed, and the weight coefficient of each dimension feature is set; According to the parameter deviation evaluation model, the Euclidean distance between the current feature vector and the standard production state feature vector is calculated and used as a preliminary deviation indicator; By using a preset proofreading module, the preliminary deviation indicator is compared with the historical deviation data to generate the preliminary evaluation result; The preliminary assessment results include the degree of deviation, the type of deviation and the trend of deviation.

4. The automatic control method for FPC production according to claim 1, characterized in that: The method of quantitatively analyzing the line accuracy and interlayer structure using a preset deviation analysis algorithm to obtain specific distribution data that causes the deviation includes: Through the deviation analysis algorithm, the line accuracy error and inter-layer structure misalignment are quantified to generate basic deviation distribution data; Based on the deviation distribution basic data, a regression analysis algorithm is used to decompose the line accuracy error data and determine the main influencing factors; Through image recognition technology, three-dimensional modeling analysis is performed on the structural dislocation data to determine the dislocation location; The dislocation positions are aligned with the main influencing factors, and basic corrections are performed through weighted fusion to obtain the specific distribution data.

5. The method according to claim 1 is an automatic control method for FPC production, characterized in that: The method of performing parameter matching calculation and designing a dynamic adjustment plan based on the specific distribution data and the current production status includes: Based on the specific distribution data, the correlation between the concentrated processing time and energy settings of the abnormal deviation points is analyzed, the processing time is prioritized according to the degree of deviation, and a time control adjustment list is formed; Combined with the current production status data, parameter matching basis is extracted from the time control adjustment list, and the parameters are classified by a support vector machine model to determine the optimal parameter combination; Extracting energy setting optimization requirements from the optimal parameter combination, generating a preliminary energy adjustment plan, and obtaining an energy optimization configuration; Analyze the dynamic adjustment execution conditions of the energy optimization configuration, formulate specific steps for dynamic adjustment, and determine the adjustment execution sequence; According to the adjustment execution sequence, the dynamic adjustment plan is obtained in combination with the control requirements and the execution plan logic mapping.

6. The automatic control method for FPC production according to claim 1, characterized in that: The updating of the processing time and energy settings in real time according to the dynamic adjustment scheme to obtain the adjusted first production configuration data includes: According to the dynamic adjustment scheme, real-time operation data is obtained from the equipment, processing time and energy settings are analyzed, and preliminary deviation assessment results are obtained; Based on the preliminary deviation assessment results, the processing time and energy settings are optimized and analyzed through a preset decision model to determine the adjustment direction and amplitude and generate preliminary control instructions; Sending the preliminary control instructions to the target device, updating the device operation status in real time and obtaining operation feedback data; If there is a deviation between the operation feedback data and the expected optimization target, a regression analysis algorithm is used to perform a secondary calibration on the processing to obtain a calibrated control instruction; Adjust production parameter configuration according to calibrated control instructions, continuously monitor equipment operating status, and obtain adjusted parameter configuration data; The first production configuration data is determined according to the parameter configuration data and the optimization target of the dynamic adjustment scheme.

7. An automatic control method for FPC production according to any one of claims 1 to 6, characterized in that: After completing the overall parameter configuration according to the first production configuration data, the method further includes: Analyze the real-time monitoring capabilities of the equipment through a preset loop feedback mechanism to determine the updated monitoring accuracy level; According to the monitoring accuracy level, the overall parameters are corrected twice using a preset deviation compensation algorithm to obtain second production configuration data; Complete the overall parameter configuration according to the second production configuration data.

8. The automatic control method for FPC production according to claim 7, characterized in that: According to the monitoring accuracy level, performing secondary correction on the overall parameters by using a preset deviation compensation algorithm to obtain second production configuration data includes: According to the monitoring accuracy level, the production parameter data is collected in real time, and pre-processed using a mean filter algorithm to obtain a smoothed parameter data set; Performing trend fitting on the parameter data set and the preset deviation standard by a linear regression algorithm to construct a deviation prediction model; Calculating compensation values of residual parameters according to the deviation prediction model to form a preliminary correction parameter set; Performing a consistency check on the preliminary calibration parameter set to obtain test data; Based on the test data, the preliminary correction parameter set is secondary corrected using a compensation value combined with a gradient descent algorithm to optimize the residual deviation to obtain an optimized parameter set; The optimized parameter set is determined as second production configuration data.

9. An automatic control device for FPC production, characterized in that: include: The data acquisition module is used to collect material characteristics and environmental conditions data in the production equipment through sensors, and pre-process the collected data to obtain analysis data streams; A standardization processing module is used to extract feature values from the analysis data stream and perform standardization processing on the extracted feature values to form a standardized feature vector set; a deviation evaluation module, configured to evaluate the parameter deviation of the production equipment based on the feature vector set to obtain a preliminary evaluation result; a deviation analysis module configured to, when the preliminary evaluation result exceeds a preset evaluation threshold, quantitatively analyze the line accuracy and interlayer structure using a preset deviation analysis algorithm to obtain specific distribution data causing the deviation; A parameter matching module is used to perform parameter matching calculations and design dynamic adjustment plans based on the specific distribution data and the current production status; A first adjustment module is configured to update the processing time and energy settings in real time according to the dynamic adjustment scheme to obtain adjusted first production configuration data; The first configuration module is used to complete overall parameter configuration according to the first production configuration data.

10. The automatic control device for FPC production according to claim 9, characterized in that: The device further comprises: A feedback adjustment module is used to analyze the real-time monitoring capabilities of the equipment through a preset loop feedback mechanism to determine the updated monitoring accuracy level; a second adjustment module, configured to perform a secondary correction on the overall parameters according to the monitoring accuracy level using a preset deviation compensation algorithm to obtain second production configuration data; The second configuration module is used to complete the overall parameter configuration according to the second production configuration data.

Images