Cyanide-free process circuit board quality tracing method and system based on artificial intelligence

Abstract

The invention discloses a cyanide-free process circuit board quality tracing method and system based on artificial intelligence, and particularly relates to the field of circuit board quality tracing, and the method comprises the steps: writing a board identifier, a hanger identifier and a process sequence number into a circuit board entering a cyanide-free process production line, extracting a fixed pixel point and a fixed electric measurement point in an edge strip region, and enabling the fixed pixel point and the fixed electric measurement point to correspond to a key process; outputting an observation index table; according to the method, the fixed observation positions of the same circuit board are continuously collected on the basis of edge calculation at each key process outlet, and the process records are formed, and then continuous comparison, abnormal undertaking judgment and same hanger merging early warning are performed on the previous and later process records according to the same observation positions. The problem that the real defect generation position of the cyanide-free process circuit board is difficult to accurately trace is solved.

Description

A Quality Traceability Method and System for Cyanide-Free Circuit Boards Based on Artificial Intelligence Technical Field

[0001] This invention relates to the field of circuit board quality traceability technology, and more specifically, to a method and system for tracing the quality of cyanide-free circuit boards based on artificial intelligence. Background Technology

[0002] In the production of cyanide-free circuit boards, existing quality control typically focuses on early defect detection and rapid retrospective investigation of related processes. A common approach involves collecting board images, electrical test results, bath parameters, equipment operating status, and batch history at pretreatment, electroplating, cleaning, post-treatment, and final inspection stages. The edge computing unit on the production line then aggregates and performs preliminary analysis, combining this data with defect detection results to trace the source of the anomaly. Taking a continuous rack plating line as an example, under conditions where multiple boards are simultaneously mounted on the same rack, continuous transfer between processes, alternating bath replenishment and process switching, significant differences in graphic loads between different boards, and limited processing time per station, localized anomalies in the preceding stages often do not immediately manifest as clear defects. Instead, they become apparent only after subsequent cleaning, reactivation, redeposition, and final inspection magnification. This can lead to… A recurring situation that can be verified on-site is that while final inspection can identify defective boards, retrospective checks often only narrow the scope to a few adjacent processes, making it difficult to further confirm the original location of the defect. Even defects with similar appearances within the same batch may correspond to different preceding states. This is because existing records mostly reflect the inspection results and final appearance of each workstation at that time, lacking a continuous record of the board's formation throughout the process flow. Consequently, subsequent analysis remains largely at the level of post-event comparison and related judgment, making it difficult to distinguish the defect's appearance location from its actual origin. Therefore, the technical problem this application aims to solve is: how to establish a quality formation record for cyanide-free process circuit boards that can be continuously retained across processes and retrospectively verified under edge computing conditions, thereby accurately determining the actual origin of the defect. Summary of the Invention

[0003] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a method and system for quality traceability of cyanide-free process circuit boards based on artificial intelligence. By continuously collecting data at fixed observation locations of the same circuit board at the exit of each key process based on edge computing and forming process records, and then performing continuous comparison, abnormal acceptance judgment, and same fixture merging warning on the records of the preceding and following processes according to the same observation locations, the problems mentioned in the background art can be solved.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a quality traceability method for cyanide-free process circuit boards based on artificial intelligence, comprising: S1, writing board identification, fixture identification, and process sequence number to the circuit boards entering the cyanide-free process production line, extracting fixed pixel points and fixed electrical test points in the edge strip area, and corresponding them with key processes, and outputting an observation index table; S2, edge computing nodes at the exit of each key process collect board surface images, local electrical test results, bath liquid parameters, and equipment parameters according to the observation index table, inputting the board surface images into a defect recognition network to obtain defect boundaries, defect categories, and defect locations, and writing them into process records according to board identification, process sequence number, and observation location, and outputting them; S3, performing coordinate matching on process records of adjacent processes under the same board identification according to the same observation location. Calculate the overlap area of ​​the defect boundary and the difference in local electrical measurement values, and classify the newly generated abnormal records, continued abnormal records, and aggravated abnormal records according to the defect category comparison results, and output the abnormal sequence; S4, scan the abnormal sequence by process number, and take the first newly generated abnormal record at the same observation position as the candidate generated record; when there are similar continued abnormal records or aggravated abnormal records in the two key processes after the candidate generated record, determine the candidate generated record as the actual defect generated record and output it; S5, merge the actual defect generated record with the process records of other circuit boards under the same fixture identification at the same process number and the same observation position; when there are two or more actual defect generated records with the same defect category and the same observation position, output the batch warning record and the corresponding process record.

[0005] In a preferred embodiment, S1 includes: S1-1, reading the edge strip area image, edge strip area layout data, and contact position data of each key process of the circuit board; establishing a unified coordinate system with the starting angle of the edge strip area as the origin and the transmission direction and board width direction; dividing the edge strip area coordinate grids with the same spacing; and outputting the edge strip area coordinate grid set; S1-2, calculating the minimum distance between the grid center and the line boundary, hole boundary, and board boundary of the edge strip area coordinate grid set grid by grid; deleting the coordinate grids that coincide with the contact position of any key process; determining the grid center of the remaining coordinate grids as fixed pixel points; and selecting a unique fixed electrical test point in the coordinate grid where each fixed pixel point is located; and outputting the observation point set; S1-3, associating the fixed pixel points and fixed electrical test points in the observation point set with the key processes according to the same coordinate grid number, and associating them with the board identification, fixture identification, and process sequence number to generate an observation index table.

[0006] In a preferred embodiment, S2 includes: S2-1, reading the plate identifier, process number, fixed pixel point and fixed electrical test point in the observation index table, acquiring the whole plate image, the local electrical test result corresponding to the fixed electrical test point, the tank liquid parameter and equipment parameter at the corresponding key process exit, and merging them according to the plate identifier and process number to output the acquisition group; S2-2, cropping local blocks from the whole plate image in the acquisition group according to the fixed pixel point coordinates, writing the center coordinates of each local block and the coordinates of the corresponding fixed electrical test point as the same observation position, and corresponding the local blocks with the local electrical test results according to the same observation position to output the observation group.

[0007] In a preferred embodiment, S2 further includes: S2-3, inputting local patches from the observation group into the defect identification network, extracting defect boundaries from the local patches, determining the defect location based on the coordinate difference between the center of the circumscribed rectangle of the defect boundary and the observation location, determining the defect category based on the image features of the area covered by the defect boundary, and outputting the identification group; S2-4, writing the defect boundaries, defect categories, and defect locations in the identification group, along with the local electrical measurement results, tank liquid parameters, and equipment parameters corresponding to the same observation location, into the process record according to the plate identifier, process number, and observation location, and outputting it.

[0008] In a preferred embodiment, S3 includes: S3-1, reading two adjacent process records under the same board identifier according to the process sequence number, extracting the defect boundary, defect category, and local electrical measurement results corresponding to the same observation position in the two process records, and forming a position comparison group according to the order of the previous process record before the next process record, and outputting the position comparison group; S3-2, performing translation on the defect boundary of the next process record in the position comparison group according to the coordinate difference between the defect position and the defect position of the previous process record, so that the translated defect boundary of the next process record and the defect boundary of the previous process record fall into the same observation position coordinate system, calculating the area of ​​the overlapping area of ​​the two as the defect boundary overlap area, and subtracting the local electrical measurement result of the previous process record from the local electrical measurement result of the next process record to obtain the local electrical measurement difference, and outputting the comparison result group.

[0009] In a preferred embodiment, S3 further includes: S3-3, comparing the comparison result group according to the defect category of the previous process record and the defect category of the next process record, recording the next process record where the defect boundary of the previous process record is empty and the defect boundary of the next process record is not empty as a new abnormal record, and recording the next process record where the defect boundary of the previous process record is not empty, the defect boundary of the next process record is not empty, the defect category is the same and the defect boundary overlap area is not zero as a continuing abnormal record, and outputting the classification result group; S3-4, calculating the defect boundary area of ​​the previous process record and the defect boundary area of ​​the next process record for the continuing abnormal records in the classification result group, rewriting the continuing abnormal records where the defect boundary area of ​​the next process record is greater than the defect boundary area of ​​the previous process record and the difference in local electrical measurement values ​​is not zero as aggravated abnormal records, and arranging the new abnormal records, continuing abnormal records and aggravated abnormal records according to the process sequence number, and outputting the abnormal sequence.

[0010] In a preferred embodiment, S4 includes: S4-1, reading abnormal records arranged by process number under the same board identifier and observation position in the abnormal sequence, extracting the newly generated abnormal record with the first process number, writing the process number, defect category, defect boundary, defect location, and local electrical test results to form a candidate generated record and outputting it; S4-2, reading the abnormal records corresponding to the first key process and the second key process after the candidate generated record, calculating the defect boundary overlap area between the two abnormal records and the candidate generated record respectively, comparing the defect categories of the two abnormal records and the candidate generated record respectively, and writing the abnormal records with the same defect category and a non-zero defect boundary overlap area as a continuation certificate record. Abnormal records with different defect categories or zero overlap area at defect boundaries are written as conflict records. The first and second process judgment results corresponding to the candidate generated records are output. S4-3: The candidate generated records whose first and second process judgment results are both renewal records are written to the pass mark. The candidate generated records whose first and second process judgment results are both renewal records are written to the review mark. The candidate generated records whose first and second process judgment results are both conflict records are written to the review mark. The candidate generated records whose first and second process judgment results are both conflict records are written to the rollback mark. The gating results are output.

[0011] In a preferred embodiment, S4 further includes: S4-4, reading the candidate generation record written with the verification mark and the process record corresponding to the same process number and the same observation position under the same fixture identifier, calculating the defect boundary overlap area between the process record and the candidate generation record and comparing the defect categories, and writing the process record with the same defect category and a non-zero defect boundary overlap area as a supporting record; when a supporting record exists, rewriting the candidate generation record as a pass record and outputting it; when no supporting record exists, rewriting the candidate generation record as a rollback record and outputting it; S4-5, reading the candidate generation record and pass record corresponding to the pass mark, writing the candidate generation record and pass record corresponding to the pass mark as the defect real generation record and outputting it, deleting the candidate generation record and rollback record corresponding to the rollback mark, and then extracting the new abnormal record with the same board identifier, the same observation position, and the process number greater than the deleted candidate generation record and the first process number, forming a new candidate generation record and re-executing the subsequent judgment; when no new candidate generation record exists, outputting the scan end mark of the current observation position.

[0012] In a preferred embodiment, S5 includes: S5-1, reading the actual defect generation record and the process record of each circuit board under the same fixture identifier, extracting the defect category, defect boundary, local electrical test result, tank liquid parameter and equipment parameter corresponding to the same process number and the same observation position, and forming a merge group according to the fixture identifier, process number and observation position, and outputting the merge group; S5-2, performing defect category comparison and observation position comparison on each actual defect generation record in each merge group, writing the actual defect generation records with the same defect category and the same observation position into the same batch record group, and counting the number of actual defect generation records in each batch record group, and outputting the batch record group; S5-3, reading the corresponding process record for each batch record group, writing the batch record group with two or more actual defect generation records into the fixture identifier, process number, observation position, defect category, corresponding board identifier and corresponding actual defect generation record, forming a batch warning record, and outputting the batch warning record and the corresponding process record.

[0013] In a preferred embodiment, an AI-based quality traceability system for cyanide-free process circuit boards includes: an index construction module, which writes board identification, fixture identification, and process sequence number to circuit boards entering the cyanide-free process production line, extracts fixed pixel points and fixed electrical test points in the edge strip area, and associates them with key processes to output an observation index table; a data acquisition module, which collects board surface images, local electrical test results, bath parameters, and equipment parameters according to the observation index table through edge computing nodes at the exit of each key process, inputs the board surface images into a defect recognition network to obtain defect boundaries, defect categories, and defect locations, and writes them into process records according to board identification, process sequence number, and observation location, and outputs them; and an anomaly determination module, which performs coordinate matching on process records of adjacent processes under the same board identification according to the same observation location to calculate... The system calculates the overlap area of ​​defect boundaries and the difference in local electrical measurement values, and classifies new abnormal records, continuing abnormal records, and aggravated abnormal records according to the defect category comparison results, outputting an abnormal sequence. The traceability and judgment module scans the abnormal sequence by process number, and takes the first new abnormal record at the same observation position as a candidate generated record. When there are similar continuing or aggravated abnormal records in the two key processes after the candidate generated record, the candidate generated record is determined as the actual defect generated record and output. The batch early warning module merges the actual defect generated record with the process records of other circuit boards under the same fixture identification at the same process number and the same observation position. When there are two or more actual defect generated records with the same defect category and the same observation position, the batch early warning record and the corresponding process record are output.

[0014] The technical effects and advantages of this invention are as follows: 1. By establishing an observation index table and continuously forming process records at the exit of each key process, anomalies of the same board can be continuously inherited across processes along the same observation position, thereby distinguishing the defect manifestation position from the actual defect generation position, and relatively improving the problem of the difficulty in accurately determining the actual generation position of defects in cyanide-free process circuit boards; 2. By writing the board image, local electrical test results, bath parameters, and equipment parameters into the same process record according to the board identification, process number, and observation position, the image results, electrical test results, and process status are closed and corresponded within the same record, reducing secondary splicing errors and record chain breaks in subsequent traceability; 3. By performing coordinate matching, defect boundary overlap area calculation, and local electrical test value difference calculation on adjacent process records, and dividing them into newly generated anomaly records, continued anomaly records, and aggravated anomaly records, the anomaly evolution process can be continuously expressed, thereby improving the temporal integrity of anomaly source judgment; 4. By using the first newly generated anomaly record as a candidate generated record, and combining the follow-up verification results of the next two key processes with the corroborating results of the records of the same fixture process, a traceability judgment is performed, which distinguishes between short-term anomalies, isolated anomalies, and continuous anomalies, and helps to suppress the situation of misjudging occasional anomalies as the actual generation location; 5. By merging the actual generation records of defects with the same process number and the same observation location under the same fixture identification, and writing the records that meet the quantity conditions into batch early warning records and simultaneously outputting the corresponding process records, the traceability of single board is further extended to the batch verification of the same fixture, thereby relatively improving the timeliness of batch anomaly detection and back-check handling. Attached Figure Description

[0015] Figure 1 is a flowchart of the method steps of the present invention.

[0016] Figure 2 is a schematic diagram of the system modules of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Referring to Figures 1-2 in the specification, the AI-based quality traceability method for cyanide-free process circuit boards of the present invention includes: S1, writing board identification, fixture identification, and process sequence number to the circuit boards entering the cyanide-free process production line, extracting fixed pixel points and fixed electrical test points in the edge strip area, and corresponding them with key processes, and outputting an observation index table; Before the circuit board enters the cyanide-free process production line, it is necessary to fix the observation positions that will be continuously used across processes, so that the board surface images and local electrical test results collected at the exits of different key processes of the same circuit board can fall on the same set of position indexes, so that process records can be continuously formed subsequently, and cross-process comparison and traceability can be performed accordingly; For this purpose, the edge strip area is first used as the layout area of ​​the observation positions, and then... The edge strip area image, edge strip area layout data, and contact position data of each key process are unified under the same coordinate reference. Within this coordinate reference, coordinate grids are divided. Then, positions affected by circuits, holes, board edges, and process contacts are eliminated. Finally, fixed pixel points and fixed electrical test points are simultaneously determined within the remaining coordinate grids, and both are written into the observation index table along with the key process and board basic identifiers. This implementation process includes the following steps: First, read the edge strip area image, edge strip area layout data, and contact position data of each key process. The edge strip area image reflects the actual appearance range of the edge strip area; the edge strip area layout data indicates the positions of circuit boundaries, hole boundaries, and board boundaries; and the contact position data indicates the positions of each key process. The sequence identifies the location of the area that has physical or conductive contact with the circuit board. After reading, the starting angle of the edge strip area is used as the origin, the transmission direction of the circuit board in the production line is used as the first coordinate direction, and the board width direction perpendicular to the transmission direction is used as the second coordinate direction. The edge strip area image and edge strip area layout data are registered to the same coordinate system. Then, the edge strip area is divided along the first and second coordinate directions at the same interval to form a coordinate grid set for the edge strip area. The same interval here uses the same side length value, which is jointly limited by the minimum line spacing, minimum hole edge distance, and fixed electrical test point layout interval in the edge strip area, and remains unchanged in one division to ensure that subsequent coordinate grid numbering, grid center calculation, and observation position mapping are all based on the same... Scale execution: After obtaining the coordinate grid set of the edge strip area, calculate the minimum distance from the grid center to the line boundary, the minimum distance to the hole boundary, and the minimum distance to the plate boundary for each coordinate grid. When any minimum distance falls within the line occupation area, hole occupation area, or outside the plate edge, delete the coordinate grid directly, so that the remaining coordinate grids are all within the observable area. Subsequently, map the contact position data of each key process to the same coordinate system, and determine whether the coordinate grid coincides with the contact position one by one. When the coordinate grid coincides with the contact position of any key process, delete the coordinate grid, so that the remaining coordinate grids are not affected by the contact of key processes. For each coordinate grid remaining after deletion, directly determine the grid center as a fixed pixel point.Then, standard electrical measurement point data is read within the coordinate grid range. Electrical measurement points falling within this grid are selected as candidate electrical measurement points, and the distance from each candidate point to the grid center is calculated. The candidate electrical measurement point with the highest distance is selected as the fixed electrical measurement point. When two candidate electrical measurement points have the same distance to the grid center, the candidate electrical measurement point with the highest number is selected according to the layout number. When there are no electrical measurement points within the coordinate grid, the coordinate grid corresponding to the fixed pixel is deleted and not added to the subsequent observation point set. After the above processing, each retained coordinate grid corresponds to only one fixed pixel and one fixed electrical measurement point, ultimately forming the observation point set. After obtaining the observation point set, the data is first processed... Each set of fixed pixels and fixed electrical measurement points in the observation point set is assigned the same coordinate grid number, and this coordinate grid number is written as the observation position, so that the fixed pixels and fixed electrical measurement points can share the same position index in subsequent processes. Then, each observation position is matched one-to-one with the key process, so that the same observation position maintains the same number and the same coordinate caliber in each key process. Then, the board identification, fixture identification, and process sequence number are written into the corresponding observation position to form an observation index table composed of board identification, fixture identification, process sequence number, observation position, fixed pixels, and fixed electrical measurement points. In this way, the board surface images subsequently collected at the exit of the key process can be... Fixed pixel point positioning allows local electrical test results to be located at fixed test points, and both can be directly mapped to the same process record through the same observation position. This ensures that subsequent process record writing, anomaly sequence formation, and defect real generation record judgment are all based on a unified location. Furthermore, by first eliminating the coordinate grids corresponding to line boundaries, hole boundaries, board boundaries, and contact positions, and then determining the fixed pixel point and fixed electrical test point, the subsequent board surface image acquisition area and local electrical test position can continuously fall within a stable area, reducing the interference of key process actions on the observation results. In practical applications: for a circuit board entering a cyanide-free process production line... First, the edge strip area image and edge strip area layout data are read. Then, the contact position data corresponding to preprocessing, electroplating, and postprocessing are read. After establishing a unified coordinate system based on the starting angle of the edge strip area, the edge strip area coordinate grid is divided. Subsequently, coordinate grids that overlap with the circuit area, hole area, outer edge area of ​​the board, and contact position are deleted. The grid center is extracted from the remaining coordinate grids as a fixed pixel point. Fixed electrical measurement points are selected in each coordinate grid according to the rule of being closest to the grid center. Finally, the fixed pixel points, fixed electrical measurement points, board identification, fixture identification, and process sequence numbers of each key process are written into the same observation index table for direct use by the edge calculation nodes at the exit of subsequent key processes.

[0019] S2. Edge computing nodes at the exits of each critical process collect panel images, local electrical measurement results, bath parameters, and equipment parameters according to the observation index table. The panel images are then input into the defect recognition network to obtain defect boundaries, defect categories, and defect locations. These are then written into the process record according to panel identification, process number, and observation location, and output. This unified collection and writing at the exit of critical processes forms the basis for subsequent process record generation, anomaly sequence construction, and defect realization. Instead of separately saving the entire panel image, local electrical measurement results, and process parameters, the observation index table is used as a reference to map fixed pixels, fixed electrical measurement points, panel identification, and process number to the same critical process exit. The collected entire panel image, local electrical measurement results, and bath parameters are then processed. The equipment parameters are aggregated into the same set of data; then, local patches are extracted from the whole board image centered on fixed pixels, and the local electrical measurement results corresponding to the local patches and fixed electrical measurement points are unified to the same observation position; then, defect identification is performed on the local patches to obtain the defect boundary, defect location, and defect category; finally, the identification results, local electrical measurement results, tank liquid parameters, and equipment parameters are written into the process record according to the board identification, process number, and observation position; after this processing, the image information, electrical measurement information, and process information at the same observation position are all included in the same process record, and secondary stitching is no longer required for subsequent cross-process comparisons; the implementation process includes the following steps: first, read the board identification, process number, fixed pixels, and fixed electrical measurement points from the observation index table. Then, at the exit of the corresponding key process, the circuit board corresponding to the board identification is collected. The whole board image is acquired by the visual acquisition unit at the exit, the local electrical test results are acquired by the electrical test unit connected to the fixed electrical test point, the tank liquid parameters are read by the online detection unit of the tank corresponding to the key process, and the equipment parameters are read by the equipment control unit corresponding to the key process. The local electrical test results are directly taken from the measured values ​​of the fixed electrical test point at the current process exit. The tank liquid parameters are read according to the temperature, conductivity, concentration and liquid level at the current process exit time. The equipment parameters are read according to the transmission speed, current value and residence time at the current process exit time. After the reading is completed, the whole board image, local electrical test results, tank liquid parameters and equipment parameters under the same board identification and the same process number are merged into a single acquisition. The data is grouped to ensure that subsequent screenshotting, recognition, and writing are all based on data from the same process moment. Then, local patches are cropped from the whole board image in the acquisition group according to fixed pixel coordinates. When cropping, the fixed pixel is used as the center, and the same length is taken on both sides along the transmission direction and the board width direction based on the side length of the coordinate grid where the fixed pixel is located, forming a rectangular area covering the current coordinate grid and its adjacent coordinate grids. This rectangular area is written as a local patch. After the local patch is generated, the coordinate grid number where the center of the local patch is located is written as the observation position. At the same time, the fixed electrical measurement point falling into the coordinate grid number is read, and the local electrical measurement result corresponding to the fixed electrical measurement point is written into the same observation position, so that the local patch and the local electrical measurement result are matched one-to-one through the same observation position.After all correspondences are completed, the local patches and local electrical measurement results corresponding to each observation position under the same board identifier and process sequence number are grouped into observation groups. Here, the observation positions still use the coordinate grid numbers in the observation index table, without changing the previous definition, thus ensuring that the observation positions in subsequent process records are completely consistent with the previous index. After obtaining the observation groups, the local patches corresponding to each observation position are input into the defect recognition network. Areas in the local patches that differ from the background area in grayscale distribution, texture direction, or edge closure are segmented, and closed defect boundaries are extracted. Then, the circumscribed rectangle of the defect boundary is calculated, and the center coordinates of the observation position are subtracted from the center coordinates of the circumscribed rectangle to obtain the defect location. In the process, the defect location is recorded using the coordinate difference relative to the center of the observation location, serving as the basis for subsequent cross-process translation and alignment. After determining the defect location, the grayscale distribution, texture density, edge direction, and color difference distribution of the defect boundary coverage area are extracted and written into the defect category according to the category number output by the defect recognition network. This allows defects of the same type to be directly compared for consistency according to the category number during subsequent process record comparisons. After all local patches are processed, an identification group is formed consisting of the defect boundary, defect location, and defect category. If a local patch does not have a closed region segmented, the defect boundary corresponding to that local patch is written as an empty boundary list, the defect category is written as an empty category marker, and the defect location is written as a zero-offset coordinate. The label is used for subsequent judgment of newly generated anomaly records; further, the defect boundary, defect category, and defect location in the identification group, along with the local electrical test results, tank liquid parameters, and equipment parameters corresponding to the same observation location, are written into the process record according to the board identification, process number, and observation location. The process record includes at least the board identification, fixture identification, process number, observation location, defect boundary, defect category, defect location, local electrical test results, tank liquid parameters, and equipment parameters. When writing, the board identification, fixture identification, process number, and observation location are written first, followed by the defect boundary, defect category, and defect location, and finally the local electrical test results, tank liquid parameters, and equipment parameters, ensuring that under the same process and observation location... Image results, electrical measurement results, and process parameters are merged into a single process record. After all process records are written, they are directly output for subsequent calculation of the overlapping area of ​​defect boundaries between adjacent processes, calculation of local electrical measurement differences, and classification of abnormal records. Through the above processing, the whole plate image, local electrical measurement results, tank liquid parameters, and equipment parameters are simultaneously acquired at the exit of the same key process. Local blocks are uniformly cropped based on fixed pixel points. Defect boundaries, defect locations, and defect categories are uniformly identified under the same observation location. The final process record has complete field sources and consistent location indexes, providing direct input for the formation of subsequent abnormal sequences and the determination of the actual generation of defect records.In practical applications: When a circuit board is transferred to the exit of the critical cyanide-free electroplating process, the edge computing node first reads the corresponding fixed pixel points and fixed electrical measurement points of the circuit board according to the observation index table. Then, it simultaneously acquires the whole image of the circuit board, the local electrical measurement results at the fixed electrical measurement points, and the temperature, conductivity, concentration, liquid level, equipment conveying speed, current value, and residence time of the cyanide-free electroplating tank. Subsequently, it extracts a local patch from the whole image according to the coordinates of the fixed pixel points, and writes the local patch and the local electrical measurement results under the same coordinate grid number as the same observation position. Then, it inputs the local patch into the defect recognition network to obtain the defect boundary, defect location, and defect category. Finally, these recognition results, along with the local electrical measurement results, tank liquid parameters, and equipment parameters, are written into the process record of the corresponding observation position and output for direct comparison between subsequent adjacent critical processes.

[0020] S3. For process records of adjacent processes under the same board identification, perform coordinate matching at the same observation position, calculate the overlapping area of ​​defect boundaries and the difference in local electrical measurement values, and classify new abnormal records, continuing abnormal records, and aggravated abnormal records according to the defect category comparison results, and output the abnormal sequence. When forming the abnormal sequence, it is necessary to continuously compare the process records formed under adjacent key processes of the same board in the same position caliber, so that the abnormalities appearing in the later process can be clearly distinguished as new, continuing, or aggravated. The focus here is not on looking at the identification results of a single process, but on matching the previous process record and the next process record one by one at the same observation position, and then performing coherent calculations on the defect boundary, defect category, and local electrical measurement results. This yields an anomaly sequence that can be directly entered into subsequent traceability judgment. To this end, firstly, two adjacent process records under the same board identifier, arranged by process number, are read to form a position comparison group. Then, the defect boundary in the later process record is shifted to the coordinate system of the observation position of the earlier process record, and the overlapping area of ​​the defect boundary and the difference in local electrical measurement values ​​are calculated. Subsequently, the later process record is classified according to whether the defect boundary is empty, whether the defect category is consistent, and whether the overlapping area of ​​the defect boundary is zero. Finally, the continuing anomaly records are further judged to determine whether they have formed aggravated anomaly records, and the anomaly sequence is written according to the process number. This implementation process includes the following steps: Firstly, two adjacent process records under the same board identifier, arranged by process number, are read, where adjacent... This refers to two consecutive process records in a critical process sequence, both of which have generated process records. If a critical process has not generated a process record, it is not considered an adjacent process for comparison. Instead, the next generated process record is read, ensuring that the position comparison is always based on two existing process records. After reading, the defect boundary, defect category, and local electrical measurement results corresponding to the same observation position are extracted from the two process records, maintaining the writing order of the preceding process record preceding the following process record. These two records are then grouped into a position comparison group. Here, the same observation position is determined by the same coordinate grid number written in the process record. As long as the observation position numbers of the preceding and following process records are consistent, they are considered to be the same observation position. Measure the position; establish separate position comparison groups for different observation positions under the same board identification, without cross-mixing; after obtaining the position comparison group, perform translation processing on the defect boundary of the next process record in the position comparison group. Specifically, first read the defect position in the previous process record and the defect position in the next process record, subtract the defect position in the previous process record from the defect position in the next process record to form a translation vector, and then subtract this translation vector from all boundary points in the defect boundary of the next process record point by point, so that the translated defect boundary of the next process record and the defect boundary of the previous process record fall into the same observation position coordinate system; then calculate the area of ​​the overlapping area enclosed by the two defect boundaries, and write the area of ​​the overlapping area as the defect boundary overlap area.The difference in local electrical measurement values ​​is obtained by subtracting the local electrical measurement result recorded in the previous process from the local electrical measurement result recorded in the subsequent process. If the local electrical measurement result is a single value, the difference is directly subtracted. If the local electrical measurement result consists of multiple measurement value fields, the difference is performed according to the main measurement value field pre-written in the process record, and the resulting difference is written as the local electrical measurement value difference. After all calculations are completed, a comparison result group containing the overlapping area of ​​the defect boundary and the local electrical measurement value difference is obtained. Then, the defect category recorded in the previous process and the defect category recorded in the subsequent process are compared in the comparison result group, and the subsequent process record is classified based on the null value status of the defect boundary recorded in the previous process and the defect boundary recorded in the subsequent process. Here, the defect boundary is a null sample. An empty boundary list is used to represent the defect category, and an empty category marker is used to indicate that the defect category is empty. When the defect boundary of the previous process record is an empty boundary list and the defect boundary of the next process record is not an empty boundary list, the next process record is recorded as a new abnormal record. When the defect boundary of the previous process record is not an empty boundary list and the defect boundary of the next process record is not an empty boundary list, the defect category of the previous process record is the same as the defect category of the next process record, and the overlap area of ​​the defect boundaries is not zero, the next process record is recorded as a continuing abnormal record. For the next process record that does not meet the above two writing conditions, neither a new abnormal record nor a continuing abnormal record is written. After completing the classification of all position comparison groups, the classification result group is output. Each classification result retains the corresponding board identifier, process number, observation location, defect boundary, defect category, defect location, and local electrical measurement result for further judgment. Furthermore, for each continuing abnormal record in the classification result group, the defect boundary area of ​​the previous process record and the defect boundary area of ​​the next process record are calculated. The defect boundary area is directly obtained from the area of ​​the region enclosed by each defect boundary. When the defect boundary area of ​​the next process record is greater than that of the previous process record, and the difference in local electrical measurement values ​​is not zero, the continuing abnormal record is rewritten as an aggravated abnormal record. When the defect boundary area of ​​the next process record is not greater than that of the previous process record, or the difference in local electrical measurement values ​​is not zero, the continuing abnormal record is rewritten as an aggravated abnormal record. When the value difference is zero, the record is kept as a continuing abnormal record and no rewriting is performed. After all rewriting is completed, the new abnormal records, continuing abnormal records, and aggravated abnormal records are arranged in ascending order of process number. When the process number is the same, they are arranged in ascending order of observation position number to form an abnormal sequence and output it. After this processing, the abnormal changes of each observation position under the same board identification in each key process can directly enter the subsequent defect real generation record judgment in a unified order. In other words, by first establishing a position comparison group with the same observation position, and then performing translation, overlap, and difference calculation, the continuous changes of defects in the preceding and following processes can be compressed into the same coordinate caliber, thereby ensuring that the abnormal sequence has a basis for repeatable calculation.In practical applications: After process records are generated for both the pretreatment and cyanide-free electroplating critical processes on a circuit board, the defect boundaries, defect categories, and local electrical measurement results with the same observation location number in these two process records are first read. Then, the defect boundaries corresponding to the cyanide-free electroplating critical process are translated according to the coordinate difference of the defect locations in the two process records, so that the two defect boundaries fall within the same observation location coordinate system, and the overlap area of ​​the defect boundaries and the difference in local electrical measurement values ​​are calculated. Subsequently, if the defect boundary at the observation location is empty in the pretreatment critical process but not empty in the cyanide-free electroplating critical process, then the cyanide-free electroplating... The process records corresponding to critical processes are recorded as newly emerging anomaly records. If the defect categories of two consecutive process records are the same and the overlap area of ​​the defect boundaries is not zero, the process record corresponding to the cyanide-free electroplating critical process is recorded as a continuing anomaly record. Furthermore, if the defect boundary area of ​​the cyanide-free electroplating critical process is greater than the defect boundary area of ​​the pretreatment critical process, and the difference in local electrical measurement values ​​is not zero, the continuing anomaly record is rewritten as an aggravated anomaly record. Finally, the newly emerging anomaly records, continuing anomaly records, and aggravated anomaly records corresponding to all observation locations on the circuit board are arranged according to the process sequence number to form an anomaly sequence for subsequent traceability and judgment.

[0021] S4. Scan the anomaly sequence by process number, and take the first newly generated anomaly record at the same observation position as a candidate generated record. When there are similar continuing anomaly records or aggravated anomaly records in the two key processes after the candidate generated record, determine the candidate generated record as the actual defect generated record and output it. After an anomaly sequence is formed at the same observation position, the earliest newly generated anomaly record cannot be directly identified as the actual defect generated record, because whether the newly generated anomaly record is subsequently continued by similar anomalies determines whether it is the previous actual generated point or just a short-term disturbance or a local misjudgment. Therefore, within the anomaly sequence of the same part identifier and the same observation position, it is necessary to first extract the candidate generated record, and then read the subsequent anomalies along the key process sequence. The system continuously assesses the overlap of defect boundaries and the consistency of defect categories between subsequent abnormal records and candidate generated records. Based on the assessment results, candidate generated records are categorized into three states: direct access, review, or regression. For candidate generated records entering the review state, process records from the same work process and observation location under the same fixture identifier are introduced as corroborating evidence to supplement the verification of whether the candidate generated record has the basis for continued retention. Finally, candidate generated records that meet the conditions for continued verification or corroborating evidence are recorded as actual defect generated records, while candidate generated records that do not meet the conditions are deleted. Newly generated abnormal records are then extracted until no new candidate generated records exist at the current observation location. Through this process, the actual defect generated records are recorded. The formation of records is no longer directly determined by a single anomaly record, but rather by the continuation of consecutive anomalies under the same board and observation position, as well as the corroborating relationship with the same fixture. This implementation process includes the following steps: First, read the anomaly records in the anomaly sequence arranged by process number under the same board identifier and observation position, and retain only the record set corresponding to that board identifier and observation position. Then, sort them by process number from smallest to largest. After sorting, find each anomaly record designated as a new anomaly record, extract the new anomaly record with the first process number, and write the process number, defect type, defect boundary, defect location, and local electrical measurement results from that record into the candidate generated record. If there are more than two new anomaly records under the same process number, count them first. Calculate the defect boundary area of ​​each newly generated anomaly record, and select the newly generated anomaly record with the largest defect boundary area as a candidate generated record; when the defect boundary areas are the same, calculate the absolute value of the local electrical measurement result corresponding to each newly generated anomaly record, and select the newly generated anomaly record with the largest absolute value as a candidate generated record; when the absolute values ​​of the local electrical measurement results are still the same, select the first record according to the writing order of the anomaly record in the anomaly sequence; after completing this process, output the candidate generated records as the starting point for subsequent determination of the current observation position; after obtaining the candidate generated records, read the anomaly records corresponding to the first key process and the second key process after the candidate generated records, where the latter first key process and the latter second key process are determined according to the key process order;If an abnormal record has been generated in the critical process immediately following the candidate generated record, then that abnormal record is used as the abnormal record corresponding to the next first critical process. If no abnormal record has been generated in the adjacent critical process, then that critical process is recorded as an empty record, and the abnormal record corresponding to the next critical process is read as the abnormal record corresponding to the next second critical process. If there are no abnormal records at the corresponding positions of either the next first or the next second critical process, then the corresponding positions are directly used as the conflict handling objects. After reading, the overlap area of ​​the defect boundaries between the two abnormal records and the candidate generated record is calculated. During the calculation, the defect boundary of the subsequent abnormal record is first translated to the coordinate system of the observation position where the candidate generated record is located according to the coordinate difference between the defect position of the candidate generated record and the defect position of the subsequent abnormal record. Next, calculate the area of ​​the overlapping region enclosed by the two defect boundaries; then, compare the defect categories of the two abnormal records and the candidate generated records respectively. Abnormal records with the same defect category and a non-zero overlapping area are written as renewal records, while abnormal records with different defect categories or zero overlapping area are written as conflict records. If there are no abnormal records in the subsequent first or second critical process, the judgment result at that position is directly written as a conflict record. After processing, output the first and second process judgment results corresponding to the candidate generated record. After obtaining the first and second process judgment results, perform cross-judgment on the same candidate generated record: when the first process judgment result is a renewal record and the second process judgment result is... When a renewal record is generated, the candidate generated record is marked with a pass-through flag. When the first process determines a renewal record and the second process determines a conflict record, the candidate generated record is marked with a review flag. When the first process determines a conflict record and the second process determines a renewal record, the candidate generated record is marked with a review flag. When the first process determines a conflict record and the second process determines a conflict record, the candidate generated record is marked with a rollback flag. After writing, a gating result is formed. The pass-through flag indicates that the candidate generated record has received the same type of abnormal acceptance in two consecutive key processes. The review flag indicates that the candidate generated record has only been accepted on one side and needs to be further verified by introducing corroborating evidence from the same fixture. The rollback flag indicates that the candidate... If a generated record is not accepted in either of the two key processes, it will no longer be retained as a current candidate generated record. For candidate generated records marked with a review mark, the process records corresponding to the same process number and the same observation position under the same fixture identifier will continue to be read, and these process records will be compared with the candidate generated records one by one. During the comparison, the defect boundaries and defect categories in the process records will be read first, and then the defect boundaries in the process records will be translated to the coordinate system of the observation position where the candidate generated record is located. Then, the overlap area of ​​the defect boundaries between the process records and the candidate generated records will be calculated, and the defect categories of the process records and the candidate generated records will be compared. When the defect categories of the process records and the candidate generated records are the same and the overlap area of ​​the defect boundaries is not zero, the process record will be written as a supporting record.After completing all supporting evidence comparisons for the same candidate generated record, check if any supporting evidence records exist. If at least one supporting evidence record exists, rewrite the candidate generated record as a pass record and output it. If no supporting evidence record exists, rewrite the candidate generated record as a rollback record and output it. The pass record retains the process number, defect category, defect boundary, defect location, and local electrical test results from the candidate generated record, which are used for subsequent direct writing into the actual defect generated record. After the gating results and verification results are formed, read the candidate generated record and pass record corresponding to the pass mark, and uniformly write the candidate generated record corresponding to the pass mark and the pass record as the actual defect generated record and output them. For the candidate generated record written to the rollback mark and the rewritten record... The candidate generated records for rollback are directly deleted from the anomaly sequence corresponding to the current observation position. After deletion, newly generated anomaly records with the same board identifier and the same observation position, whose process number is greater than the deleted candidate generated record and whose process number is the first, are reread and formed into new candidate generated records. Then, the subsequent anomaly record reading, renewal record and conflict record writing, gating result writing, and corresponding processing of circumstantial verification are re-executed in sequence. When there are no new newly generated anomaly records with the same board identifier and the same observation position, the current board identifier and the current observation position are written to the scan end marker and output to end the subsequent extraction and judgment of that observation position. After this processing, multiple newly generated anomaly records under the same observation position can be processed sequentially. The judgment process continues until a true defect generation record is retained or all scans are completed. Through the above processing, the selection rules for newly generated anomaly records under the same board identifier and the same observation position, the reading rules for the first and second key processes, the continued verification judgment rules for the overlapping area of ​​defect boundaries, the writing rules for through markers, verification markers, and rollback markers, the supplementary verification rules for supporting evidence records, and the rescanning rules after rollback are all fixed. This prevents subsequent issues such as the inability to uniquely determine candidate generation records, the inability to handle missing subsequent anomaly records, the lack of supporting evidence for verification records, the inability to continue after rollback, or the inability to terminate scanning at the current observation position. Furthermore, the succession relationship between the last two key processes is used to solidify the candidate generation records. Then, by supplementing and verifying the process records of the same fixture, short-term anomalies, isolated anomalies, and continuous anomalies can be distinguished, so that the final defect generation record has a continuous basis and a basis for corroboration of the same fixture. In practical application: when three anomaly records are generated sequentially at the same observation position of a circuit board, the first newly generated anomaly record is extracted as a candidate generation record according to the process sequence number. Then, the anomaly records corresponding to the first and second key processes are read, and the defect boundary overlap area between these two anomaly records and the candidate generation record is calculated and the defect category is compared. If the two subsequent anomaly records are consistent with the defect category of the candidate generation record and the defect boundary overlap area is not zero, then the candidate generation record is directly written as the defect generation record.If only one subsequent abnormal record meets the conditions, then further read the process records corresponding to the same process and observation position under the same fixture identifier, and check if there is any corroborating record with the same defect category as the candidate generated record and whose defect boundary overlap area is not zero. If it exists, then rewrite the candidate generated record as a passed record and write it as the actual defect generated record. If it does not exist, then delete the candidate generated record, and then extract the next newly generated abnormal record and continue to perform the same judgment. When there are no new newly generated abnormal records under the current observation position, output the scan end marker for that observation position.

[0022] S5. Merge the actual defect generation record with the process records of other circuit boards under the same fixture identifier at the same process number and observation position. When there are two or more actual defect generation records with the same defect category and the same observation position, output a batch warning record and the corresponding process record. After completing the determination of the actual defect generation record, it is necessary to further determine whether there is a co-occurrence of multiple boards with the same generation characteristics under the same fixture identifier, so as to extend the traceability of the single board to the batch warning. The focus here is no longer on the individual actual defect generation record itself, but on putting the process records and actual defect generation records corresponding to the same fixture identifier, the same process number, and the same observation position into the same merging range, and then processing the actual defect generation records within the merging range. The generated records are grouped and counted according to their similarity. Finally, groups that meet the quantity criteria are written as batch warning records, and these batch warning records are output together with the corresponding process records that constitute the group. After this processing, the batch warning records retain both the concentrated characteristics of batch anomalies and the traceability chain that can be traced back to the original process records. The implementation process includes the following steps: First, read the actual defect generation records and the process records of each circuit board under the same fixture identifier. The actual defect generation records serve as the master records for merging and participate in subsequent grouping, while the process records of each circuit board serve as the base records for merging and provide the corresponding process information. During reading, first filter out all records under the same fixture according to the fixture identifier, and then extract the defect categories corresponding to the same process number and the same observation position. The data, including the defect type, defect boundary, local electrical measurement results, tank liquid parameters, and equipment parameters, are written together with the corresponding part identifier into the same merging unit. Each merging unit corresponds to only one fixture identifier, one process number, and one observation location. If multiple process numbers or observation locations exist under the same fixture identifier, they are formed into different merging units, without cross-mixing. After all merging units are written, they are grouped into merging groups and output. After obtaining the merging groups, defect category comparison and observation location comparison are performed on each actual defect generation record in each merging group. Here, the observation location comparison is still based on the observation location number written in the process record, and the defect category comparison is still based on the category number output by the defect identification network. When two actual defect generation records are... When the defect categories and observation locations are consistent, these two actual defect generation records are written into the same batch record group. If there are three or more actual defect generation records in a certain merge group, the first actual defect generation record is used as the benchmark, and each subsequent record is compared with the defect category and observation location. Actual defect generation records that meet the writing conditions are continuously merged into the same batch record group, while actual defect generation records that do not meet the writing conditions are formed into a new batch record group. After the processing of all merge groups is completed, the number of actual defect generation records in each batch record group is counted. When counting, different board identifiers are used as the counting unit, and actual defect generation records with the same board identifier are counted only once. After the statistics are completed, the batch record group is output.After forming batch record groups, the corresponding process records for each batch record group are read, and batch record groups with two or more actual defect generation records are selected. For batch record groups that meet this quantity condition, the fixture identifier, process number, observation position, defect category, corresponding board identifier, and corresponding actual defect generation record are written into the same batch warning record. The corresponding board identifier is written as the set of all board identifiers participating in the counting within the batch record group, and the corresponding actual defect generation record is written as the set of all actual defect generation records participating in the counting within the batch record group. After the batch warning record is formed, the process records corresponding to the batch record group are extracted, and the batch warning record and the corresponding process record are output together. Here, the corresponding process record refers to all process records that constitute the basis for merging the batch record group, used for subsequent back-checking of the original observation situation under the same fixture with the same process number and the same observation position, thus ensuring that the batch warning record can directly fall back to the original record chain. Through the above processing, the hierarchical relationship between actual defect generation records, merging groups, batch record groups, and batch warning records is fixed, and batch anomalies under the same fixture identifier no longer rely on general... The judgment of abnormal quantity is not based on the number of abnormal items, but on the continuous merging of items with the same process number, the same observation position, and the same defect category. At the same time, the batch warning record is output synchronously with the corresponding process record, so that the original process record constituting the batch warning can be directly reviewed during subsequent verification. In practical application: when multiple circuit boards under the same fixture have completed the judgment of the actual defect generation record, these actual defect generation records and the process records of all circuit boards under the fixture are read first. Then, records with the same process number and the same observation position are merged into the same merging group. Subsequently, within each merging group, the defect category and observation position of the actual defect generation record are compared one by one. The actual defect generation records with the same defect category and the same observation position are written into the same batch record group, and the number of actual defect generation records in each batch record group is counted according to different board identifiers. Finally, for batch record groups with two or more items, the fixture identifier, process number, observation position, defect category, corresponding board identifier set, and corresponding actual defect generation record set are written to form a batch warning record. The batch warning record is output together with the process record corresponding to the merging group for the production line to perform subsequent batch verification and handling.

[0023] Furthermore, the AI-based quality traceability system for cyanide-free process circuit boards includes: an index construction module, which writes board identification, fixture identification, and process sequence number to the circuit boards entering the cyanide-free process production line, extracts fixed pixels and fixed electrical test points in the edge strip area, and associates them with key processes to output an observation index table; a data acquisition module, which collects board surface images, local electrical test results, bath parameters, and equipment parameters according to the observation index table through edge computing nodes at the exit of each key process, inputs the board surface images into a defect recognition network to obtain defect boundaries, defect categories, and defect locations, and writes them into process records according to board identification, process sequence number, and observation location, and outputs them; and an anomaly judgment module, which judges the anomalies by analyzing the process records of adjacent processes under the same board identification. Coordinate matching is performed at the same observation location to calculate the overlapping area of ​​defect boundaries and the difference in local electrical measurement values. Based on the defect category comparison results, newly generated abnormal records, continuing abnormal records, and aggravated abnormal records are classified, and anomaly sequences are output. The traceability and judgment module scans the abnormal sequence by process number, identifying the first newly generated abnormal record at the same observation location as a candidate generated record. When a continuing or aggravated abnormal record of the same type exists in the two key processes following the candidate generated record, the candidate generated record is determined as the actual defect generated record and output. The batch early warning module merges the actual defect generated record with process records of other circuit boards under the same fixture identifier at the same process number and observation location. When there are two or more defect categories... When defects at the same observation location are actually generated and records are generated, the working principle of outputting batch warning records and corresponding process records is as follows: This scheme first predetermines a set of reusable observation locations in the edge strip area of ​​the circuit board, and writes the fixed pixel point, fixed electrical test point, board identification, fixture identification and process number corresponding to each observation location into the observation index table; then, at the exit of each key process, the edge computing node synchronously collects the board surface image, local electrical test results, tank liquid parameters and equipment parameters according to this index table, and writes them into process records; next, the process records of the same board in adjacent key processes are compared according to the same observation location, the overlap area of ​​the defect boundary and the difference of local electrical test values ​​are calculated, and newly generated abnormal records and continued abnormal records are identified in turn. The process involves recording and aggravating abnormal records to form an abnormal sequence. Candidate generated records are then extracted one by one from this sequence, and combined with subsequent verification in key processes and corroborating evidence from records on the same fixture, to screen out the truly valid defect-generated records that can be continuously accepted. Finally, the truly valid defect-generated records under the same fixture, with the same process number and the same observation location are merged to form batch warning records and corresponding process records. The preceding indexing ensures consistent locations, the subsequent continuous comparison ensures traceable abnormal evolution, and the subsequent verification and corroborating evidence judgment ensures that the true generation location can be confirmed. Therefore, the entire process does not only consider the final inspection results but also connects the entire chain of defects from their appearance and continuation to their mass spread along the technological process.For example, on a continuous production line for cyanide-free circuit boards, the same rack carries multiple circuit boards through key processes such as pretreatment, cyanide-free electroplating, and post-treatment. Edge computing nodes are deployed at the exit of each key process. These edge computing nodes collect local images and electrical measurement data for each board at fixed observation positions according to a pre-written observation index table, and simultaneously read the bath parameters and equipment parameters for that process. If a board shows no obvious abnormalities after pretreatment, but a defect boundary first appears at a certain observation position after cyanide-free electroplating, and the same defect persists at that position after post-treatment with the boundary continuing to expand, then the system will identify the corresponding cyanide-free electroplating line. The abnormal record serves as a candidate generation record. Combined with the verification results of subsequent processes and corroborating records from other boards in the same location on the same rack, it is confirmed that the defect first formed during the cyanide-free electroplating process. If two or more circuit boards on the same rack generate similar defects at the same process number and observation location, the system will output corresponding batch warning records and process records, reminding the production line to promptly check the bath solution status and equipment status corresponding to that process. Thus, in practical applications, on-site personnel no longer only see which board is faulty, but can directly see in which critical process and observation location the defect began to form, and whether it has already spread throughout the same rack.

[0024] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for quality traceability of cyanide-free process circuit boards based on artificial intelligence, characterized in that, include: S1. Write board identification, fixture identification, and process number to the circuit boards entering the cyanide-free process production line. Extract fixed pixels and fixed electrical test points in the edge strip area and match them with key processes, outputting an observation index table. S2. At the exit of each key process, edge computing nodes collect board images, local electrical test results, bath parameters, and equipment parameters according to the observation index table. Input the board images into the defect recognition network to obtain defect boundaries, defect categories, and defect locations. Write these into process records according to board identification, process number, and observation location, and output them. S3. Perform coordinate matching on process records of adjacent processes under the same board identification at the same observation location. Calculate the overlapping area of ​​defect boundaries and the difference in local electrical test values. Classify new abnormal records, continuing abnormal records, and aggravated abnormal records according to the defect category comparison results, and output the abnormal sequence. S4. Scan the abnormal sequence by process number. Take the first new abnormal record at the same observation location as a candidate generated record. When there are similar continuing or aggravated abnormal records in the two key processes following the candidate generated record, determine the candidate generated record as the actual defect generated record and output it. S5. Merge the actual defect generation record with the process records of other circuit boards under the same fixture identifier at the same process number and the same observation position; when there are two or more actual defect generation records with the same defect category and the same observation position, output the batch warning record and the corresponding process record.

2. The method for quality traceability of cyanide-free process circuit boards based on artificial intelligence according to claim 1, characterized in that: S1 includes: S1-1, reading the edge strip area image, edge strip area layout data, and contact position data of each key process of the circuit board; establishing a unified coordinate system with the starting angle of the edge strip area as the origin and the transmission direction and board width direction; dividing the edge strip area coordinate grids with the same spacing; and outputting the edge strip area coordinate grid set; S1-2, calculating the minimum distance between the grid center and the line boundary, hole boundary, and board boundary of the edge strip area coordinate grid set grid by grid; deleting the coordinate grids that coincide with the contact position of any key process; determining the grid center of the remaining coordinate grids as fixed pixel points; and selecting a unique fixed electrical test point in the coordinate grid of each fixed pixel point; and outputting the observation point set; S1-3, associating the fixed pixel points and fixed electrical test points in the observation point set with the key processes according to the same coordinate grid number, and associating them with the board identification, fixture identification, and process sequence number to generate an observation index table.

3. The method for quality traceability of cyanide-free process circuit boards based on artificial intelligence according to claim 2, characterized in that: S2 includes: S2-1, reading the plate identifier, process number, fixed pixel point and fixed electrical test point from the observation index table, collecting the whole plate image, the local electrical test result corresponding to the fixed electrical test point, the tank liquid parameter and equipment parameter at the corresponding key process exit, and merging them according to the plate identifier and process number to output the acquisition group; S2-2, cropping local blocks from the whole plate image in the acquisition group according to the fixed pixel point coordinates, writing the center coordinates of each local block and the coordinates of the corresponding fixed electrical test point as the same observation position, and corresponding the local blocks with the local electrical test results according to the same observation position to output the observation group.

4. The method for quality traceability of cyanide-free process circuit boards based on artificial intelligence according to claim 3, characterized in that: S2 further includes: S2-3, inputting the local map pieces in the observation group into the defect recognition network, extracting the defect boundaries in the local map pieces, determining the defect location according to the coordinate difference between the center of the outer rectangle of the defect boundary and the observation location, and determining the defect category according to the image features of the area covered by the defect boundary, and outputting the recognition group; S2-4, writing the defect boundaries, defect categories, and defect locations in the recognition group, along with the local electrical test results, tank liquid parameters, and equipment parameters corresponding to the same observation location, into the process record according to the plate identification, process number, and observation location, and outputting it.

5. The method for quality traceability of cyanide-free process circuit boards based on artificial intelligence according to claim 4, characterized in that: S3 includes: S3-1, reading two adjacent process records under the same board identifier according to the process sequence number, extracting the defect boundary, defect type and local electrical measurement results corresponding to the same observation position in the two process records, and forming a position comparison group according to the order of the previous process record before the next process record, and outputting the position comparison group; S3-2, performing translation on the defect boundary of the next process record in the position comparison group according to the coordinate difference between the defect position and the defect position of the previous process record, so that the translated defect boundary of the next process record and the defect boundary of the previous process record fall into the same observation position coordinate system, calculating the area of ​​the overlapping area of ​​the two as the defect boundary overlap area, and subtracting the local electrical measurement result of the previous process record from the local electrical measurement result of the next process record to obtain the local electrical measurement difference, and outputting the comparison result group.

6. The method for quality traceability of cyanide-free process circuit boards based on artificial intelligence according to claim 5, characterized in that: S3 further includes: S3-3, comparing the defect categories of the previous process record and the defect categories of the next process record in the comparison result group, recording the next process record where the defect boundary of the previous process record is empty and the defect boundary of the next process record is not empty as a new abnormal record, and recording the next process record where the defect boundary of the previous process record is not empty, the defect boundary of the next process record is not empty, the defect category is the same and the defect boundary overlap area is not zero as a continuing abnormal record, and outputting the classification result group; S3-4, calculating the defect boundary area of ​​the previous process record and the defect boundary area of ​​the next process record for the continuing abnormal records in the classification result group, rewriting the continuing abnormal records where the defect boundary area of ​​the next process record is greater than the defect boundary area of ​​the previous process record and the difference in local electrical measurement values ​​is not zero as aggravated abnormal records, and arranging the new abnormal records, continuing abnormal records and aggravated abnormal records according to the process sequence number, and outputting the abnormal sequence.

7. The method for quality traceability of cyanide-free process circuit boards based on artificial intelligence according to claim 6, characterized in that: S4 includes: S4-1, reading abnormal records arranged by process number under the same board identifier and observation position in the abnormal sequence, extracting the newly generated abnormal record with the first process number, writing the process number, defect type, defect boundary, defect location, and local electrical test results to form a candidate generated record and outputting it; S4-2, reading the abnormal records corresponding to the first and second key processes after the candidate generated records, calculating the defect boundary overlap area between the two abnormal records and the candidate generated records respectively, comparing the defect types of the two abnormal records and the candidate generated records respectively, writing the abnormal records with the same defect type and a non-zero defect boundary overlap area as continued verification records, and writing the abnormal records with different defect types as continued verification records. Alternatively, abnormal records with zero overlap area at the defect boundary are written as conflict records, and the first and second process judgment results corresponding to the candidate generated records are output; S4-3, write the candidate generated records whose first and second process judgment results are both renewal records to the pass mark, write the candidate generated records whose first and second process judgment results are both renewal records to the review mark, write the candidate generated records whose first and second process judgment results are both conflict records to the review mark, write the candidate generated records whose first and second process judgment results are both conflict records to the rollback mark, and output the gating result.

8. The method for quality traceability of cyanide-free process circuit boards based on artificial intelligence according to claim 7, characterized in that: S4 further includes: S4-4, reading the candidate generation record written with the verification mark and the process record corresponding to the same process number and the same observation position under the same fixture identifier, calculating the defect boundary overlap area between the process record and the candidate generation record and comparing the defect categories, and writing the process record with the same defect category and a non-zero defect boundary overlap area as a supporting record; when a supporting record exists, rewriting the candidate generation record as a pass record and outputting it; when no supporting record exists, rewriting the candidate generation record as a rollback record and outputting it; S4-5, reading the candidate generation record and pass record corresponding to the through mark, writing the candidate generation record and pass record corresponding to the through mark as the defect real generation record and outputting it, deleting the candidate generation record and rollback record corresponding to the rollback mark, and then extracting the new abnormal record with the same board identifier, the same observation position, and the process number greater than the deleted candidate generation record and the first process number, forming a new candidate generation record and re-executing the subsequent judgment; when no new candidate generation record exists, outputting the scan end mark of the current observation position.

9. The method for quality traceability of cyanide-free process circuit boards based on artificial intelligence according to claim 8, characterized in that: S5 includes: S5-1, reading the actual defect generation record and the process record of each circuit board under the same fixture identifier, extracting the defect category, defect boundary, local electrical test result, tank liquid parameter and equipment parameter corresponding to the same process number and the same observation position, and forming a merge group according to the fixture identifier, process number and observation position, and outputting the merge group; S5-2, performing defect category comparison and observation position comparison on each actual defect generation record in each merge group, writing the actual defect generation records with the same defect category and the same observation position into the same batch record group, and counting the number of actual defect generation records in each batch record group, and outputting the batch record group; S5-3, reading the corresponding process record for each batch record group, writing the batch record group with two or more actual defect generation records into the fixture identifier, process number, observation position, defect category, corresponding board identifier and corresponding actual defect generation record, forming a batch warning record, and outputting the batch warning record and the corresponding process record.

10. A quality traceability system for cyanide-free process circuit boards based on artificial intelligence, characterized in that: include: The index building module writes board identifiers, fixture identifiers and process numbers to the circuit boards entering the cyanide-free process production line, extracts fixed pixel points and fixed electrical test points in the edge strip area, and associates them with key processes to output an observation index table. The data acquisition module collects plate images, local electrical measurement results, tank liquid parameters and equipment parameters through edge computing nodes at the exit of each key process according to the observation index table. The plate images are input into the defect recognition network to obtain the defect boundary, defect category and defect location. The data is then written into the process record and output according to the plate identification, process number and observation location. The anomaly determination module performs coordinate matching on the process records of adjacent processes under the same board part identification according to the same observation position, calculates the overlapping area of ​​the defect boundary and the difference of local electrical measurement values, and classifies new anomaly records, continued anomaly records and aggravated anomaly records according to the defect category comparison results, and outputs the anomaly sequence. The traceability and judgment module scans the abnormal sequence by process number and takes the first newly generated abnormal record at the same observation position as a candidate generated record. When there are similar continuing abnormal records or aggravated abnormal records in the two key processes after the candidate generated record, the candidate generated record is determined to be the actual defect generated record and output. The batch warning module merges the actual defect generation records with the process records of other circuit boards under the same fixture identification at the same process number and the same observation position. When there are two or more actual defect generation records with the same defect category and the same observation position, it outputs the batch warning record and the corresponding process record.

Images