Design method for checking interlayer dislocation of multilayer circuit board

Abstract

The invention provides a design method for checking dislocation between layers of a multilayer circuit board, which comprises the following steps of: designing an intelligent positioning identifier at a board corner of each layer, capturing an SPM high-definition image, transmitting the image to a central control system through a high-speed network, designing special image processing software to automatically identify the SPM, and extracting concentric circle and rectangular bar code information; the method comprises the steps of calculating relative position deviation between layers, preprocessing original image data, converting the original image data into structured offset data, establishing a real-time data analysis platform, generating a detailed interlayer offset report, setting a threshold value alarm mechanism, providing a visual interface to display the state and historical trend of a production line, and obtaining a real-time data analysis result based on real-time data analysis. A machine learning algorithm is adopted to predict an offset trend, subsequent process parameters are optimized, interlayer offset in a final finished product is reduced, a dynamic process control system is introduced, operation parameters of production equipment are automatically adjusted according to real-time feedback, and production stability and precision are ensured.

Description

Technical Field

[0001] This invention relates to the field of printed circuit board manufacturing technology, and in particular to a design method for inspecting interlayer misalignment in multilayer circuit boards. Background Technology

[0002] Multilayer circuit boards play a crucial role in modern electronic devices and are widely used in fields such as computers, communication equipment, automotive electronics, and medical devices. As electronic products develop towards miniaturization, high performance, and multifunctionality, higher requirements are placed on the manufacturing precision and reliability of multilayer circuit boards, especially the interlayer alignment precision, which is directly related to the electrical performance and reliability of the products.

[0003] Traditionally, interlayer offset detection of multilayer circuit boards mainly relies on manual measurement and visual inspection. This method is not only time-consuming and labor-intensive, but also easily affected by human factors, leading to inconsistent detection results. Especially in large-scale production environments, manual inspection cannot meet the needs of efficient production. Traditional image processing methods usually use pixel-level precision to calculate interlayer offset, with a minimum resolution of a few pixels. For products with high precision requirements, this calculation method cannot meet the needs and is prone to missing tiny offset problems.

[0004] Most existing detection systems use a static threshold alarm mechanism, which means that a fixed offset threshold is preset. When the detected offset exceeds the threshold, an alarm is triggered. This method lacks flexibility and cannot be dynamically adjusted according to different production processes and product requirements, and is prone to false alarms or missed alarms. Summary of the Invention

[0005] The purpose of this invention is to provide a design method for checking interlayer misalignment in multilayer circuit boards.

[0006] The problem this invention aims to solve is the challenge of detecting and correcting interlayer misalignment during the production of multilayer circuit boards. By designing intelligent positioning markers (SPMs) at the corners of each layer and combining them with high-resolution optical cameras, dedicated image processing software, and a real-time data analysis platform, the invention enables precise measurement and dynamic adjustment of interlayer misalignment. By introducing machine learning algorithms to predict misalignment trends, the invention optimizes production process parameters and establishes a closed-loop control system. Through centralized database and big data analysis, the invention identifies potential problem patterns, proposes improvement suggestions, and formulates preventative maintenance plans.

[0007] A design method for detecting interlayer misalignment in multilayer circuit boards, the technical solution of which is as follows: S1: Design an intelligent positioning mark (SPM) at the corner of each layer. The SPM consists of multiple concentric circles and a set of rectangular barcodes. The concentric circles are used to determine the interlayer offset, while the rectangular barcodes contain coded information, including the layer number and production batch. The center point of each SPM serves as the reference origin for that layer. High-contrast metal is used as the material selection for the SPM. S2: At the nodes of the production line, including after inner layer etching, after lamination and riveting, and after lamination and target hitting, a high-resolution optical camera is installed to capture high-definition images of the SPM and transmit them to the central control system via the network. Dedicated image processing software is developed to automatically identify the SPM, extract information from concentric circles and rectangular barcodes, and calculate the relative positional deviation between each layer. S3: Preprocess the raw image data acquired by the automated inspection system, convert it into structured offset data, establish a real-time data analysis platform, summarize and analyze the data from each inspection point, generate detailed inter-layer offset reports, set a threshold alarm mechanism, and immediately trigger an alarm when the detected offset exceeds the set standard. Provide operators with an intuitive visualization interface to display the current status of the production line, the offset of each layer, and historical trend charts. S4: Based on the results of real-time data analysis, machine learning algorithms are used to predict possible offset trends and optimize the parameters of subsequent processes accordingly. The pressing pressure, heating time and cooling rate are adjusted according to the known offset patterns to reduce the interlayer offset in the final product. A dynamic process control system is introduced to automatically adjust the operating parameters of the production equipment according to the real-time feedback data, including automatically slowing down the production line speed when a large offset is found in a certain layer, giving more time for fine calibration. S5: Store all test data in a centralized database to form a quality record archive, save detailed information for each batch of products, including comparative data before and after each adjustment, regularly perform statistical analysis on the data in the database, identify potential problem patterns and regularities, use big data technology and artificial intelligence algorithms to uncover factors affecting interlayer offset, and propose targeted improvement suggestions. Based on the data analysis results, formulate preventive maintenance plans, replace vulnerable parts in advance, and adjust equipment status.

[0008] Furthermore, in S1, the concentric circles are used to determine the interlayer offset, while the rectangular barcode contains encoding information, including the layer number and production batch. The center point of each SPM serves as the reference origin for that layer, including: S11: Set up SPM at the four corners of each layer board to arrange an all-round detection network. Design multiple concentric circles at the corner of each layer board. The diameter of each circle is different, and the diameter of the circle increases by 6mil for each layer. The width of the circle line is 3mil. S12: Integrate rectangular barcodes into SPM, containing basic information such as the number of layers and production batch, as well as process parameter codes, including temperature, pressure, and etching time; S13: Set the center point of each SPM as the reference origin of that layer, and perform dynamic calibration through an automated detection system. If the reference origin of a certain layer is found to be offset during the production process, the parameters of subsequent processes will be automatically adjusted.

[0009] Furthermore, a high-resolution optical camera is installed in S2 to extract information from the concentric circles and rectangular barcodes, and to calculate the relative positional deviation between each layer, including: S21: Install high-resolution optical cameras at nodes in the production line, including after inner layer etching, lamination and riveting, and lamination and target assembly. Calibrate the camera position at each node, especially on multi-layer circuit boards with complex shapes and curved surfaces. Equip each camera with an environmental adaptive lighting system. S22: Edge computing technology is introduced at the camera end to perform preliminary processing and compression on the acquired images, and the images are transmitted to the central control system via high-speed industrial Ethernet. S23: Develop dedicated image processing software to automatically identify concentric circle and rectangular barcodes and extract information from them. Apply deep learning algorithms to train and optimize images, and adapt to different types of multilayer circuit boards and production processes through a large amount of historical data and simulated scenarios. S24: Employs a sub-pixel precision calculation method to measure the relative positional deviation between layers.

[0010] Furthermore, in step S3, the raw image data acquired by the automated detection system is preprocessed, and a threshold alarm mechanism is set, including: S31: A multi-stage noise removal algorithm is adopted, which removes high-frequency noise through a low-pass filter, eliminates random noise through an adaptive median filter, enhances the preprocessed image, and automatically extracts the features of SPM, including the center of the concentric circles and the position of the rectangular barcode, using a deep learning model. S32: Multi-dimensional fusion of offset data from different detection nodes to form a complete three-dimensional offset map, and real-time data stream processing technology to update the offset data in real time during the production process; S33: Establish a real-time data analysis platform based on a distributed computing architecture to detect anomalies in offset data, identify inter-layer offsets that exceed the normal range, set a dynamic threshold alarm mechanism, and dynamically adjust the alarm threshold according to different production processes and product requirements.

[0011] Furthermore, in step S4, machine learning algorithms are used to predict possible deviation trends, and the operating parameters of the production equipment are automatically adjusted based on real-time feedback data, including: S41: A time series analysis model is used to model historical offset data, feature engineering techniques are used to extract key features from the offset data, and pattern recognition algorithms are combined to identify different types of offset patterns. S42: Employing a reinforcement learning algorithm, the system simulates different pressing pressure settings to evaluate their impact on the final product offset, thereby finding the optimal pressing pressure configuration. S43: Dynamically adjust heating time and cooling rate based on real-time offset data, and use a multi-objective optimization algorithm to combine the impact of multiple process parameters, including pressing pressure, heating time, and cooling rate, on product quality. S44: Establish a real-time feedback closed-loop control system to feed back the offset data collected by the sensors to the control system in real time. When a large offset is detected in a certain layer, the system will automatically slow down the production line speed to allow time for fine calibration. The system adopts an intelligent scheduling algorithm to dynamically adjust the operating sequence of production equipment and resource allocation based on the real-time offset data.

[0012] The beneficial effects of this invention are as follows: It adopts a sub-pixel-level precision calculation method, with a minimum resolution of 0.1 micrometers, which can accurately measure minute interlayer shifts. Through multi-stage noise removal and image enhancement processing, it ensures the consistency and stability of image quality, reduces detection errors caused by environmental changes, and applies deep learning algorithms to train and optimize images, thereby improving the accuracy and speed of SPM feature extraction and adapting to different types of multilayer circuit boards and manufacturing processes. A real-time feedback closed-loop control system is established to feed back the offset data collected by sensors to the control system in real time, so as to adjust the operating parameters of the production equipment in a timely manner, reduce downtime and rework rate in the production process, apply intelligent scheduling algorithm, dynamically adjust the operating sequence and resource allocation of production equipment according to real-time offset data, optimize the production process, improve the overall efficiency of the production line, and reduce manual intervention through automated detection system and dynamic calibration mechanism, thereby improving the automation level of the production process and reducing the workload of operators. Attached Figure Description

[0013] Figure 1 This is a flowchart of a design method for checking interlayer misalignment in multilayer circuit boards. Detailed Implementation

[0014] The present invention will be further described clearly and completely below, but the scope of protection of the present invention is not limited thereto.

[0015] A design method for detecting interlayer misalignment in multilayer circuit boards, the technical solution of which is as follows: S1: Design an intelligent positioning mark (SPM) at the corner of each layer. The SPM consists of multiple concentric circles and a set of rectangular barcodes. The concentric circles are used to determine the interlayer offset, while the rectangular barcodes contain coded information, including the layer number and production batch. The center point of each SPM serves as the reference origin for that layer. High-contrast metal is used as the material selection for the SPM. S2: At the nodes of the production line, including after inner layer etching, after lamination and riveting, and after lamination and target hitting, a high-resolution optical camera is installed to capture high-definition images of the SPM and transmit them to the central control system via the network. Dedicated image processing software is developed to automatically identify the SPM, extract information from concentric circles and rectangular barcodes, and calculate the relative positional deviation between each layer. S3: Preprocess the raw image data acquired by the automated inspection system, convert it into structured offset data, establish a real-time data analysis platform, summarize and analyze the data from each inspection point, generate detailed inter-layer offset reports, set a threshold alarm mechanism, and immediately trigger an alarm when the detected offset exceeds the set standard. Provide operators with an intuitive visualization interface to display the current status of the production line, the offset of each layer, and historical trend charts. S4: Based on the results of real-time data analysis, machine learning algorithms are used to predict possible offset trends and optimize the parameters of subsequent processes accordingly. The pressing pressure, heating time and cooling rate are adjusted according to the known offset patterns to reduce the interlayer offset in the final product. A dynamic process control system is introduced to automatically adjust the operating parameters of the production equipment according to the real-time feedback data, including automatically slowing down the production line speed when a large offset is found in a certain layer, giving more time for fine calibration. S5: Store all test data in a centralized database to form a quality record archive, save detailed information for each batch of products, including comparative data before and after each adjustment, regularly perform statistical analysis on the data in the database, identify potential problem patterns and regularities, use big data technology and artificial intelligence algorithms to uncover factors affecting interlayer offset, and propose targeted improvement suggestions. Based on the data analysis results, formulate preventive maintenance plans, replace vulnerable parts in advance, and adjust equipment status.

[0016] refer to Figure 1 The diagram shown is a flowchart of a design method for checking interlayer misalignment in multilayer circuit boards.

[0017] Furthermore, in S1, the concentric circles are used to determine the interlayer offset, while the rectangular barcode contains encoding information, including the layer number and production batch. The center point of each SPM serves as the reference origin for that layer, including: S11: Set up SPM at the four corners of each layer board to arrange an all-round detection network. Design multiple concentric circles at the corner of each layer board. The diameter of each circle is different, and the diameter of the circle increases by 6mil for each layer. The width of the circle line is 3mil. S12: Integrate rectangular barcodes into SPM, containing basic information such as the number of layers and production batch, as well as process parameter codes, including temperature, pressure, and etching time; S13: Set the center point of each SPM as the reference origin of that layer, and perform dynamic calibration through an automated detection system. If the reference origin of a certain layer is found to be offset during the production process, the parameters of subsequent processes will be automatically adjusted.

[0018] Furthermore, a high-resolution optical camera is installed in S2 to extract information from the concentric circles and rectangular barcodes, and to calculate the relative positional deviation between each layer, including: S21: Install high-resolution optical cameras at nodes in the production line, including after inner layer etching, lamination and riveting, and lamination and target assembly. Calibrate the camera position at each node, especially on multi-layer circuit boards with complex shapes and curved surfaces. Equip each camera with an environmental adaptive lighting system. S22: Edge computing technology is introduced at the camera end to perform preliminary processing and compression on the acquired images, and the images are transmitted to the central control system via high-speed industrial Ethernet. S23: Develop dedicated image processing software to automatically identify concentric circle and rectangular barcodes and extract information from them. Apply deep learning algorithms to train and optimize images, and adapt to different types of multilayer circuit boards and production processes through a large amount of historical data and simulated scenarios. S24: Employs a sub-pixel precision calculation method to measure the relative positional deviation between layers.

[0019] Furthermore, in step S3, the raw image data acquired by the automated detection system is preprocessed, and a threshold alarm mechanism is set, including: S31: A multi-stage noise removal algorithm is adopted, which removes high-frequency noise through a low-pass filter, eliminates random noise through an adaptive median filter, enhances the preprocessed image, and automatically extracts the features of SPM, including the center of the concentric circles and the position of the rectangular barcode, using a deep learning model. S32: Multi-dimensional fusion of offset data from different detection nodes to form a complete three-dimensional offset map, and real-time data stream processing technology to update the offset data in real time during the production process; S33: Establish a real-time data analysis platform based on a distributed computing architecture to detect anomalies in offset data, identify inter-layer offsets that exceed the normal range, set a dynamic threshold alarm mechanism, and dynamically adjust the alarm threshold according to different production processes and product requirements.

[0020] Furthermore, in step S4, machine learning algorithms are used to predict possible deviation trends, and the operating parameters of the production equipment are automatically adjusted based on real-time feedback data, including: S41: A time series analysis model is used to model historical offset data, feature engineering techniques are used to extract key features from the offset data, and pattern recognition algorithms are combined to identify different types of offset patterns. S42: Employing a reinforcement learning algorithm, the system simulates different pressing pressure settings to evaluate their impact on the final product offset, thereby finding the optimal pressing pressure configuration. S43: Dynamically adjust heating time and cooling rate based on real-time offset data, and use a multi-objective optimization algorithm to combine the impact of multiple process parameters, including pressing pressure, heating time, and cooling rate, on product quality. S44: Establish a real-time feedback closed-loop control system to feed back the offset data collected by the sensors to the control system in real time. When a large offset is detected in a certain layer, the system will automatically slow down the production line speed to allow time for fine calibration. The system adopts an intelligent scheduling algorithm to dynamically adjust the operating sequence of production equipment and resource allocation based on the real-time offset data.

[0021] This invention provides a design method for inspecting interlayer misalignment in multilayer circuit boards. It involves designing intelligent positioning markers at the corners of each layer, capturing high-definition images of surface-mount particulate matter (SPM), and transmitting them to a central control system via a high-speed network. Dedicated image processing software automatically identifies the SPM, extracts concentric circles and rectangular barcode information, calculates the relative positional deviation between layers, preprocesses the raw image data to convert it into structured offset data, establishes a real-time data analysis platform, generates detailed interlayer offset reports, sets threshold alarm mechanisms, and provides an intuitive visual interface to display production line status and historical trends. Based on real-time data analysis, machine learning algorithms predict offset trends, optimize subsequent process parameters, reduce interlayer offset in the final product, and introduce a dynamic process control system that automatically adjusts the operating parameters of production equipment based on real-time feedback to ensure production stability and accuracy.

Claims

1. A design method for inspecting interlayer misalignment in multilayer circuit boards, characterized in that, include: S1: Design an intelligent positioning mark (SPM) at the corner of each layer. The SPM consists of multiple concentric circles and a set of rectangular barcodes. The concentric circles are used to determine the interlayer offset, while the rectangular barcodes contain coded information, including the layer number and production batch. The center point of each SPM serves as the reference origin for that layer. High-contrast metal is used as the material selection for the SPM. S2: At the nodes of the production line, including after inner layer etching, after lamination and riveting, and after lamination and target hitting, a high-resolution optical camera is installed to capture high-definition images of the SPM and transmit them to the central control system via the network. Dedicated image processing software is developed to automatically identify the SPM, extract information from concentric circles and rectangular barcodes, and calculate the relative positional deviation between each layer. S3: Preprocess the raw image data acquired by the automated inspection system, convert it into structured offset data, establish a real-time data analysis platform, summarize and analyze the data from each inspection point, generate detailed inter-layer offset reports, set a threshold alarm mechanism, and immediately trigger an alarm when the detected offset exceeds the set standard. Provide operators with an intuitive visualization interface to display the current status of the production line, the offset of each layer, and historical trend charts. S4: Based on the results of real-time data analysis, machine learning algorithms are used to predict possible offset trends and optimize the parameters of subsequent processes accordingly. The pressing pressure, heating time and cooling rate are adjusted according to the known offset patterns to reduce the interlayer offset in the final product. A dynamic process control system is introduced to automatically adjust the operating parameters of the production equipment according to the real-time feedback data, including automatically slowing down the production line speed when a large offset is found in a certain layer, giving more time for fine calibration. S5: Store all test data in a centralized database to form a quality record archive, save detailed information for each batch of products, including comparative data before and after each adjustment, regularly perform statistical analysis on the data in the database, identify potential problem patterns and regularities, use big data technology and artificial intelligence algorithms to uncover factors affecting interlayer offset, and propose targeted improvement suggestions. Based on the data analysis results, formulate preventive maintenance plans, replace vulnerable parts in advance, and adjust equipment status.

2. The design method for inspecting interlayer misalignment in a multilayer circuit board as described in claim 1, characterized in that, In S1, concentric circles are used to determine interlayer offset, while rectangular barcodes contain encoded information, including layer number and production batch. The center point of each SPM serves as the reference origin for that layer, including: S11: Set up SPM at the four corners of each layer board to arrange an all-round detection network. Design multiple concentric circles at the corner of each layer board. The diameter of each circle is different, and the diameter of the circle increases by 6mil for each layer. The width of the circle line is 3mil. S12: Integrate rectangular barcodes into SPM, containing basic information such as the number of layers and production batch, as well as process parameter codes, including temperature, pressure, and etching time; S13: Set the center point of each SPM as the reference origin of that layer, and perform dynamic calibration through an automated detection system. If the reference origin of a certain layer is found to be offset during the production process, the parameters of subsequent processes will be automatically adjusted.

3. The design method for inspecting interlayer misalignment in a multilayer circuit board as described in claim 1, characterized in that, A high-resolution optical camera is installed in S2 to extract information from concentric circle and rectangular barcodes, and to calculate the relative positional deviation between each layer, including: S21: Install high-resolution optical cameras at nodes in the production line, including after inner layer etching, lamination and riveting, and lamination and target assembly. Calibrate the camera position at each node, especially on multi-layer circuit boards with complex shapes and curved surfaces. Equip each camera with an environmental adaptive lighting system. S22: Edge computing technology is introduced at the camera end to perform preliminary processing and compression on the acquired images, and the images are transmitted to the central control system via high-speed industrial Ethernet. S23: Develop dedicated image processing software to automatically identify concentric circle and rectangular barcodes and extract information from them. Apply deep learning algorithms to train and optimize images, and adapt to different types of multilayer circuit boards and production processes through a large amount of historical data and simulated scenarios. S24: Employs a sub-pixel precision calculation method to measure the relative positional deviation between layers.

4. The design method for inspecting interlayer misalignment in a multilayer circuit board as described in claim 1, characterized in that, S3 involves preprocessing the raw image data acquired by the automated detection system and setting a threshold alarm mechanism, including: S31: A multi-stage noise removal algorithm is adopted, which removes high-frequency noise through a low-pass filter, eliminates random noise through an adaptive median filter, enhances the preprocessed image, and automatically extracts the features of SPM, including the center of the concentric circles and the position of the rectangular barcode, using a deep learning model. S32: Multi-dimensional fusion of offset data from different detection nodes to form a complete three-dimensional offset map, and real-time data stream processing technology to update the offset data in real time during the production process; S33: Establish a real-time data analysis platform based on a distributed computing architecture to detect anomalies in offset data, identify inter-layer offsets that exceed the normal range, set a dynamic threshold alarm mechanism, and dynamically adjust the alarm threshold according to different production processes and product requirements.

5. The design method for inspecting interlayer misalignment in a multilayer circuit board as described in claim 1, characterized in that, The S4 process employs machine learning algorithms to predict potential deviation trends and automatically adjusts the operating parameters of the production equipment based on real-time feedback data, including: S41: A time series analysis model is used to model historical offset data, feature engineering techniques are used to extract key features from the offset data, and pattern recognition algorithms are combined to identify different types of offset patterns. S42: Employing a reinforcement learning algorithm, the system simulates different pressing pressure settings to evaluate their impact on the final product offset, thereby finding the optimal pressing pressure configuration. S43: Dynamically adjust heating time and cooling rate based on real-time offset data, and use a multi-objective optimization algorithm to combine the impact of multiple process parameters, including pressing pressure, heating time, and cooling rate, on product quality. S44: Establish a real-time feedback closed-loop control system to feed back the offset data collected by the sensors to the control system in real time. When a large offset is detected in a certain layer, the system will automatically slow down the production line speed to allow time for fine calibration. The system adopts an intelligent scheduling algorithm to dynamically adjust the operating sequence of production equipment and resource allocation based on the real-time offset data.

Images