An AI-based fully automatic copper thickness measuring instrument for PCB boards

Abstract

The present invention belongs to the technical field of copper plating thickness measurement, and discloses a fully automatic copper thickness measuring instrument for PCB boards based on artificial intelligence. The present invention locates and identifies the area of the coating to be detected, and then adaptively locates the copper thickness detection points. This analysis method focuses on the key parts and improves the detection efficiency; it distributes points flexibly according to the coating distribution, comprehensively collects data, improves the detection accuracy, reduces the omission error, and effectively guarantees the reliability of product quality detection. The present invention performs data cleaning on the tilt angle and corrects the copper thickness cosine based on the copper thickness detection correction model. This analysis method can remove abnormal tilt angle data, ensure the reliability of the measurement angle, and at the same time can accurately adjust the data according to the influence of the probe tilt on the copper thickness measurement. The combination of the two improves the accuracy of copper thickness measurement, effectively reduces the measurement error, ensures that the measurement result truly reflects the copper thickness of the PCB board, and improves the credibility of product quality detection.

Description

Technical Field

[0001] The present invention belongs to the technical field of copper plating thickness measurement, and relates to a fully automatic copper thickness measuring instrument for PCB boards based on artificial intelligence. Background Art

[0002] The PCB board is a core component of electronic devices, used for mechanically supporting and electrically connecting electronic components. The PCB board usually consists of a substrate, a copper layer, a solder mask layer, and a silk screen layer. Among them, the copper plating layer of the PCB board can provide a highly conductive circuit path to ensure the integrity of signal transmission. The thickness of the copper plating layer directly affects the current-carrying capacity, signal integrity, structural reliability, and manufacturing cost of the PCB. Therefore, the research on the fully automatic copper thickness measurement of PCB boards is of great significance.

[0003] In the prior art, there are also technical solutions for measuring the thickness of copper plating layers. For example, a Chinese invention patent application for an on-line double-sided copper thickness tester with the publication number CN108827170A, which consists of a tester, a feeding bracket, and a discharging bracket. The feeding bracket and the discharging bracket are respectively located on both sides of the tester, and a conveying wheel group is provided at the top. The test conveying wheel group in the middle of the tester corresponds to them. There are vertical slide rails on both sides of the test conveying wheel group, and upper and lower test devices are equipped. This tester can simultaneously perform copper thickness tests on the copper plating layers on both sides of the circuit board. During the test, the device moves along the slide rail to complete a comprehensive inspection. This design ensures product quality, improves production efficiency, realizes high automation, and can also reduce production costs.

[0004] In addition, a Chinese invention patent application for a method for quickly calculating the copper thickness on the PCB surface with the publication number CN115164803A includes obtaining a qualitative curve representing the relationship between different copper thicknesses and PAD widths using a surface copper thickness gauge. Then, based on these curves, combined with the input PAD width and the display value of the copper thickness gauge, the actual copper thickness of the PCB to be measured is deduced, and the corresponding relationship between the PAD width, the display value of the measuring instrument, and the actual copper thickness is recorded to form an analog comparison table. During actual measurement, the operator only needs to look up the analog comparison table to quickly obtain the actual copper thickness.

[0005] Although the above two solutions propose some solutions for measuring the thickness of copper plating layers, there are still certain limitations. For example, on the one hand, taking the above on-line double-sided copper thickness tester as an example, although this design scheme proposes a method for double-sided thickness measurement, it does not further adaptively select the thickness acquisition points. In terms of measurement accuracy, it does not consider the copper thickness distribution difference and the complex circuit board structure, and it is easy to miss abnormal areas and difficult to reflect the actual situation of the copper thickness of each layer; in terms of production adaptability, it lacks flexibility and is difficult to adapt to the changes in circuit board specifications and personalized production requirements, increasing the debugging cost.

[0006] On the other hand, taking the above-mentioned method for quickly calculating the copper thickness on the PCB surface as an example, although this design scheme proposes to construct a mapping relationship model between the PAD width and the actual copper thickness, it lacks a comprehensive copper thickness detection for cleaning the measured copper thickness data and correcting the copper thickness cosine, and cannot eliminate the abnormal data that may exist in the measurement, affecting the measurement accuracy; it is difficult to adjust the errors caused by factors such as the tilt of the measurement probe. Summary of the Invention

[0007] In view of this, to solve the problems raised in the above-mentioned background technology, a fully automatic PCB copper thickness measuring instrument based on artificial intelligence is proposed.

[0008] The object of the present invention can be achieved by the following technical solutions: A fully automatic PCB copper thickness measuring instrument based on artificial intelligence, comprising: a transmission and positioning module, which is provided with an automatic transmission device and a mobile detection device. The automatic transmission device is composed of a front-end transmission unit and a rear-end transmission unit, and the mobile detection device is composed of a copper thickness detection probe and a control system.

[0009] A fixed-point selection module, which uses the front-end transmission unit to transfer the target PCB board to the monitoring area, uses an image acquisition device to collect the surface image of the target PCB board, locates and identifies the area of the coating to be detected, and adaptively locates the copper thickness detection points at the same time.

[0010] A copper thickness measurement module, which uses the copper thickness detection probe to move to each copper thickness detection point to perform multiple copper thickness detections to obtain a number of collected copper thicknesses. At the same time, it uses an attitude sensor to obtain the tilt angle and direction of the copper thickness detection probe, cleans the data of the tilt angle to obtain an effective tilt angle, and then inputs it into a pre-set copper thickness detection correction model and combines it with the collected copper thickness to perform copper thickness cosine correction to obtain the monitored copper thickness of each copper thickness detection point.

[0011] A defect identification module, which judges whether the target PCB board has abnormalities based on the monitored copper thickness, and identifies and outputs the abnormal area.

[0012] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) By locating and identifying the area of the coating to be detected, and then adaptively locating the copper thickness detection points, this analysis method focuses on the key parts and improves the detection efficiency; flexible point layout according to the coating distribution, comprehensive data collection, improved detection accuracy, reduced omission errors, and effectively guarantees the reliability of product quality detection.

[0013] (2) By cleaning the tilt angle data and performing copper thickness cosine correction based on the copper thickness detection correction model in the present invention, this analysis method can remove abnormal tilt angle data, ensure the reliability of the measured angle, and accurately adjust the data according to the influence of the probe tilt on the copper thickness measurement. The combination of the two improves the accuracy of copper thickness measurement, effectively reduces measurement errors, ensures that the measurement results truly reflect the copper thickness of the PCB board, and enhances the credibility of product quality inspection. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for describing the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0015] Figure 1 Schematic diagram of the connection of each module of the system of the present invention.

[0016] Figure 2 Schematic diagram of the structure of a full-automatic copper thickness measuring instrument provided by the present invention.

[0017] Figure 3 Schematic diagram of the detection principle of the copper thickness detection probe of a full-automatic copper thickness measuring instrument provided by the present invention.

[0018] Figure 4 Schematic diagram of the distribution of copper thickness detection points corresponding to an embodiment provided by the present invention using the five-point detection method.

[0019] Reference numerals: 1 - front-end conveying unit, 2 - rear-end conveying unit, 3 - copper thickness detection probe, 4 - control system, 5 - target PCB board, 6 - upper surface copper thickness detection probe, 7 - lower surface copper thickness detection probe, 8 - upper surface of the target PCB board, 9 - lower surface of the target PCB board. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only some of the embodiments of the present invention, rather than all of them. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts belong to the scope of protection of the present invention.

[0021] Please refer to Figure 1As shown in the figure, the present invention provides a fully automatic copper thickness measuring instrument for PCB boards based on artificial intelligence, which includes a transmission and positioning module, a fixed-point selection module, a copper thickness measurement module, and a defect recognition module. Among them, the transmission and positioning module is connected to the fixed-point selection module, the fixed-point selection module is connected to the copper thickness measurement module, and the copper thickness measurement module is connected to the defect recognition module.

[0022] Please refer to Figure 2 As shown in the figure, the transmission and positioning module is used to set an automatic transmission device and a mobile detection device. The automatic transmission device is composed of a front-end transmission unit and a rear-end transmission unit, and the mobile detection device is composed of a copper thickness detection probe and a control system.

[0023] It should be explained that the automatic transmission device is composed of a front-end transmission unit and a rear-end transmission unit. In the whole measurement process, the front-end transmission unit is responsible for transporting the target PCB board from the initial position to the monitoring area, laying the foundation for subsequent detection work; the rear-end transmission unit will transport the PCB board to the subsequent process or the designated position after the detection is completed. This way of cooperation between the front and rear ends realizes the automatic transmission of the PCB board in the measuring instrument, improves the detection efficiency, and reduces manual intervention.

[0024] It should be explained that the mobile detection device is composed of a copper thickness detection probe and a control system. The copper thickness detection probe is a key component directly used to measure the copper thickness of the PCB board. It can move to each point to be detected under the command of the control system to collect copper thickness data. The control system is responsible for precisely controlling the movement of the copper thickness detection probe, including the movement path, speed, and detection time point, etc., to ensure that the copper thickness detection probe can complete the detection task accurately and efficiently.

[0025] Please refer to Figure 3 As shown in the figure, in a preferred embodiment of the present invention, the copper thickness detection probe is composed of an upper surface copper thickness detection probe and a lower surface copper thickness detection probe. The upper surface copper thickness detection probe is used to monitor the copper thickness of the upper surface of the target PCB board, and the lower surface copper thickness detection probe is used to monitor the copper thickness of the lower surface of the target PCB board.

[0026] It should be noted that in the present invention, for the convenience of function description and explanation, the corresponding functions and implementation methods are not specifically described separately for the upper and lower surfaces. However, in practice, the principles, functions, implementation methods, and related processes of copper thickness detection on the upper and lower surfaces are the same.

[0027] The fixed-point selection module is used to transport the target PCB board to the monitoring area by using the front-end transmission unit, collect the surface image of the target PCB board by using an image acquisition device, locate and identify the plating area to be detected, and at the same time adaptively locate the copper thickness detection points.

[0028] In a preferred embodiment of the present invention, the specific method for identifying the area of the coating to be detected is as follows: Obtain the chromaticity values of each pixel point on the surface image of the target PCB board, and then compare them with the pre-set chromaticity values of the copper coating to match the set of pixel points with consistent chromaticity values. Then, record the continuous pixel points as the same coating area.

[0029] It should be noted that in the selected set of pixel points, continuous pixel points are grouped into the same coating area. In the image, if some pixel points are adjacent to each other in spatial position without interruption, they are considered continuous. The basis for this is that in an actual PCB board, the copper coating is usually continuously distributed and there will be no irregular dispersion between pixel points. In this way, the copper coating area on the entire surface image of the PCB board can be divided, laying a foundation for further determining the area of the coating to be detected.

[0030] Perform a coincidence test on each coating area and the pre-set target coating shape to obtain the area of the coincidence area, and then conduct a matching analysis to obtain the similarity between each coating area and the target coating shape. Combine with the pre-set similarity threshold to compare and determine whether each coating area is the area of the coating to be detected.

[0031] It should be added that the specific method for analyzing the similarity between each coating area and the target coating shape: Calculate the ratio of the area of the coincidence area to the area of each coating area to obtain the similarity.

[0032] It should be noted that the copper thickness detection of the present invention only targets the coating of the circuit pattern and does not involve the detection of the copper thickness of the hole wall.

[0033] Record the coating area with a similarity greater than the pre-set similarity threshold as the area of the coating to be detected.

[0034] It should be added that a comparison is made in combination with the pre-set similarity threshold. The similarity threshold is determined according to actual production experience and detection accuracy requirements, and it is a judgment criterion. Compare the similarity value of each coating area with the target coating shape with this threshold. If the similarity of a certain coating area is greater than the pre-set similarity threshold, it is considered that this coating area is the area of the coating to be detected. On the contrary, if the similarity is less than the threshold, it indicates that there may be a large deviation in this coating area and it does not meet the detection requirements, and it is not recognized as the area of the coating to be detected. In this way, the areas that truly need to be detected for copper thickness can be screened out, improving the pertinence and efficiency of the detection.

[0035] In a preferred embodiment of the present invention, the specific method for adaptively positioning the copper thickness detection points is as follows: Obtain the center points of each plating area to be detected, denoted as the position monitoring points corresponding to the plating areas to be detected. Connect the position monitoring points of each plating area to be detected with the position monitoring points of other plating areas to be detected, and obtain the connection lengths as the monitoring distribution distances between each plating area to be detected and other plating areas to be detected.

[0036] Compare the monitoring distribution distances with the adaptively distributed distance thresholds obtained through pre-analysis. Regard the plating areas to be detected corresponding to the monitoring distribution distances less than the adaptively distributed distance thresholds as suspected associated areas. Count the number of suspected associated areas corresponding to each plating area to be detected, and then select the plating area to be detected corresponding to the minimum number of suspected associated areas as the initial monitoring plating area. Then, perform iterative identification of the monitoring plating areas to obtain each monitoring plating area of the target PCB board.

[0037] It should be explained that each monitoring distribution distance is compared with this threshold. Those plating areas to be detected with monitoring distribution distances less than the adaptively distributed distance thresholds are identified as suspected associated areas. This is because the plating areas with smaller distances are physically closer to each other. There may be mutual influences or similar characteristics between them, so they are regarded as suspected associated areas.

[0038] It should be explained that for each plating area to be detected, count the number of suspected associated areas corresponding to it. Among the numerous plating areas to be detected, select the plating area to be detected with the minimum number of suspected associated areas as the initial monitoring plating area. This is because the minimum number means that this area is relatively less affected by other areas and has stronger independence. Starting the detection from this area can obtain its copper thickness data more purely, avoiding being interfered by too many complex factors at the initial stage of detection, and ensuring the accuracy and stability of the detection results.

[0039] Please refer to Figure 4 As shown, position five copper thickness detection points of each monitoring plating area according to the preset five-point detection method, and summarize to obtain the copper thickness detection points of the target PCB board.

[0040] It should be explained that the preset five-point detection method is a detection point positioning method that is commonly used in the industry or specifically formulated according to the detection requirements of this measuring instrument. Take one point at the center of each monitoring plating area, and then take one point at each of its four vertices or the midpoints of its four sides. By setting detection points at multiple key positions, the overall situation of the copper thickness in this plating area can be comprehensively reflected, avoiding missing possible copper thickness differences within the area due to only detecting a single point.

[0041] It should be noted that all monitored plating areas are summarized according to the copper thickness detection points determined by the five-point detection method. There are usually multiple monitored plating areas on a PCB board. The five detection points of each area are summarized to form the copper thickness detection point set of the entire target PCB board. These detection points cover all the plating areas that need to be detected on the PCB board. By detecting the copper thickness at these detection points, comprehensive and accurate copper thickness data can be obtained, providing sufficient basis for subsequent judgment of whether there are problems such as abnormal copper thickness on the PCB board.

[0042] In a preferred embodiment of the present invention, the specific method for the adaptive distribution spacing threshold is as follows: Obtain the distances between the position monitoring points of each plating area to be detected and the position monitoring points of other plating areas to be detected, and then compare and select the minimum distance as the minimum interval distance of each plating area to be detected. Then, perform an average calculation to obtain the adaptive distribution spacing threshold.

[0043] It should be noted that the minimum distance is selected because it represents the closest interval situation between the plating area to be detected and other plating areas, reflecting the closest connection degree between this area and the surrounding plating areas. In this way, a minimum interval distance is determined for each plating area to be detected. These minimum interval distances are important basic data for subsequent calculation of the adaptive distribution spacing threshold. When determining the adaptive distribution spacing threshold, by performing an average calculation on the minimum interval distances of all plating areas to be detected, a value that comprehensively reflects the general interval situation between plating areas can be obtained. This value is the adaptive distribution spacing threshold. It is used to judge the correlation degree between plating areas and thus plays an important role in links such as adaptively locating copper thickness detection points.

[0044] It should be noted that the present invention locates and identifies the plating areas to be detected, and then adaptively locates the copper thickness detection points. This analysis method focuses on key parts and improves the detection efficiency; it distributes points flexibly according to the plating distribution, comprehensively collects data, improves the detection accuracy, reduces omission errors, and effectively guarantees the reliability of product quality detection.

[0045] The copper thickness measurement module is used to move the copper thickness detection probe to each copper thickness detection point to perform multiple copper thickness detections to obtain a number of collected copper thicknesses. At the same time, use the attitude sensor to obtain the tilt angle and direction of the copper thickness detection probe, perform data cleaning on the tilt angle to obtain the effective tilt angle, and then input it into a pre-set copper thickness detection correction model and jointly perform copper thickness cosine correction with the collected copper thickness to obtain the monitored copper thickness of each copper thickness detection point.

[0046] It should be noted that during the copper thickness detection process, the inclination of the copper thickness detection probe is closely related to the unevenness of the copper layer. Since the copper layer on the PCB board is not absolutely flat and there are unevennesses. When the copper thickness detection probe touches the protruding part of the copper layer, the reaction force received by the probe is uneven, resulting in the probe being unable to maintain a vertical and stable state and thus tilting; when encountering a sunken area of the copper layer, the probe also needs to adjust its angle in order to be able to touch the sunken area for detection, thus resulting in tilting.

[0047] In a preferred embodiment of the present invention, the specific method for the effective inclination angle is as follows: The inclination angles and directions of the copper thickness detection probes in each copper thickness detection process are respectively compared differently to obtain the inclination angle difference index and direction deviation index between each copper thickness detection and other copper thickness detection processes.

[0048] It should be added that the inclination angle refers to the included angle of the copper thickness detection probe relative to the vertical direction, and the direction deviation refers to the included angle of the copper thickness detection probe on the upper surface relative to the preset reference direction in the horizontal direction.

[0049] It should be added that the specific analysis method for the inclination angle difference index: The inclination angles of the copper thickness detection probes in other copper thickness detection processes are averaged to obtain the reference inclination angle, and then the inclination angle of the copper thickness detection probe in the current copper thickness detection process is subtracted from the reference inclination angle and then the ratio is calculated with the reference inclination angle to obtain the inclination angle difference index.

[0050] The specific analysis method for the direction deviation index refers to the specific analysis method for the inclination angle difference index.

[0051] The inclination angle difference index and the direction deviation index are respectively compared with the preset inclination angle difference index threshold and direction deviation index threshold to determine whether the results of each copper thickness detection are valid.

[0052] It should be noted that the preset inclination angle difference index threshold and direction deviation index threshold are determined according to the accuracy requirements of the measuring instrument, the detection environment, and a large amount of experimental data. These thresholds represent the acceptable range of changes in the inclination angle and direction of the copper thickness detection probe under normal detection conditions. Under specific measuring instruments and detection conditions, through multiple tests, it is found that when the inclination angle difference index does not exceed a certain value and the direction deviation index is within a certain range, the detection results are less affected by the probe posture and can ensure higher accuracy, and this range value is set as the threshold.

[0053] It should be added that if both the inclination angle difference index and the direction deviation index are less than the preset inclination angle difference index threshold and direction deviation index threshold, then it is recognized that the result of this copper thickness detection is valid.

[0054] The tilt angle obtained from the copper thickness detection with a valid judgment result is recorded as the valid tilt angle.

[0055] In a preferred embodiment of the present invention, the specific construction method of the copper thickness detection correction model is as follows: Use a copper thickness detection probe to perform a finite number of copper thickness detections on several plating areas of several PCB boards to obtain the corresponding copper thickness. At the same time, obtain the tilt angle of the copper thickness detection probe during each copper thickness detection process. Perform pre-correction on each copper thickness and the corresponding tilt angle of the copper thickness detection probe to obtain the monitored copper thickness of each copper thickness detection, and perform deviation analysis with the corresponding copper thickness to obtain the copper thickness deviation degree.

[0056] Associate the tilt angle of the copper thickness detection probe with the copper thickness deviation degree. Then, calculate the average value of the copper thickness deviation degrees corresponding to the same tilt angle of the copper thickness detection probe to obtain the copper thickness deviation degree mapped by each tilt angle of the copper thickness detection probe. Then, summarize to obtain the mapping relationship between the tilt angle of the copper thickness detection probe - copper thickness deviation degree, and use it as the copper thickness detection correction model.

[0057] It should be explained that the reasons for the above model construction are as follows: First, improve the detection accuracy. By associating the tilt angle with the copper thickness deviation degree, the measurement error caused by the tilt of the probe can be corrected, making the measurement result closer to the true copper thickness. Second, have good adaptability and can cope with different tilt situations of the probe in various complex detection environments, and find the corresponding deviation degree according to the model for correction. Finally, improve the detection efficiency. Take the average value of the copper thickness deviation degrees corresponding to the same tilt angle, simplify the data processing process, can quickly find the mapping relationship during a large number of detections, efficiently obtain the correction data, and achieve the rapid batch detection of the copper thickness of PCB boards.

[0058] In a preferred embodiment of the present invention, the specific method of copper thickness cosine correction is as follows: Input the valid tilt angle into the copper thickness detection correction model to obtain the corresponding copper thickness deviation degree.

[0059] Perform a product calculation on the copper thickness deviation degree and the collected copper thickness to obtain the copper thickness correction amount, and then sum it with the collected copper thickness to obtain the monitored copper thickness of each copper thickness detection point.

[0060] The defect recognition module is used to judge whether there is an abnormality in the target PCB board based on the monitored copper thickness, identify and output the abnormal area.

[0061] In a preferred embodiment of the present invention, the specific analysis process for judging whether there is an abnormality in the target PCB board is as follows: Perform deviation analysis on the monitored copper thickness of each copper thickness detection point and the preset standard copper thickness to obtain the monitored copper thickness deviation degree of each copper thickness detection point, and then compare it with the preset monitored copper thickness deviation degree threshold to identify whether there is a copper thickness abnormality at each copper thickness detection point.

[0062] Statistically identify the number of copper thickness detection points with abnormal copper thickness, and then calculate the proportion of the total number of copper thickness detection points to obtain the proportion of the number of abnormal copper thickness detection points.

[0063] Compare the proportion of the number of abnormal copper thickness detection points with the preset threshold of the proportion of the number of abnormal copper thickness detection points to determine whether the target PCB board is abnormal.

[0064] It should be added that if the proportion of the number of abnormal copper thickness detection points is greater than the threshold of the proportion of the number of abnormal copper thickness detection points, it is identified that the target PCB board is abnormal; otherwise, it is identified that the target PCB board is not abnormal.

[0065] It should be noted that determining the threshold of the proportion of the number of abnormal copper thickness detection points requires considering multiple factors. First, according to the PCB board quality standard, different-purpose PCB boards have different requirements for copper thickness. The threshold for high-precision PCB boards is low, while the threshold for ordinary PCB boards is high. Second, refer to the production process capabilities. A mature and stable process can set a lower threshold, while a process with large fluctuations requires a higher threshold. It is also possible to conduct statistical analysis based on a large amount of historical data, collect copper thickness detection data, study the relationship between the abnormal proportion and the product qualification rate and defective rate, and find a value that can balance quality and efficiency.

[0066] In a preferred embodiment of the present invention, the specific method for identifying and outputting the abnormal area is as follows: Locate the positions of each abnormal copper thickness detection point, divide the target PCB board into several grid areas according to the preset grid size, and then calculate the proportion of abnormal copper thickness detection points in each grid area.

[0067] Sort the proportions of abnormal copper thickness detection points in each grid area from largest to smallest, and record the grid area with the largest proportion of abnormal copper thickness detection points as the abnormal area and output it.

[0068] It should be noted that the purpose of identifying and outputting the abnormal area: 1. Accurately locate the problem area: By locating the positions of each abnormal copper thickness detection point and dividing the grid area to calculate the abnormal proportion, the area with the most serious copper thickness abnormality on the PCB board can be accurately found. For example, in a complex PCB board circuit layout, the specific small area with copper thickness problems can be accurately locked, rather than vaguely judging the situation of the whole board, which is convenient for subsequent targeted processing.

[0069] 2. Improve the production quality control ability: After quickly determining the abnormal area, production personnel can analyze and process these areas in a timely manner, such as adjusting production process parameters, checking relevant production links, etc., to avoid the performance problems of the PCB board caused by abnormal copper thickness from flowing into the next process, and effectively improve the overall quality of the product.

[0070] 3. Facilitate fault analysis and optimization: After identifying the abnormal area, it helps to analyze the faults in the production process. For example, if the copper thickness anomaly frequently occurs in a specific area, the production equipment components, raw material supply and other links corresponding to this area can be inspected key points, and then the entire production process can be optimized to improve the stability and reliability of production.

[0071] It should be noted that the present invention performs data cleaning on the tilt angle and corrects the copper thickness cosine based on the copper thickness detection correction model. This analysis method can remove abnormal tilt angle data, ensure the reliability of the measured angle, and at the same time can accurately adjust the data according to the influence of the probe tilt on the copper thickness measurement. The combination of the two improves the accuracy of the copper thickness measurement, effectively reduces the measurement error, ensures that the measurement result truly reflects the copper thickness of the PCB board, and improves the credibility of the product quality inspection.

[0072] The above content is only an example and explanation of the concept of the present invention. Those skilled in the art of the present technology make various modifications or supplements or use similar methods to replace the specific embodiments described, as long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, they should all belong to the protection scope of the present invention.

Claims

1. An artificial intelligence-based fully automatic PCB board copper thickness measuring instrument, characterized in that, Including: A transmission and positioning module, which sets an automatic transmission device and a movement detection device. The automatic transmission device consists of a front-end transmission unit and a rear-end transmission unit, and the movement detection device consists of a copper thickness detection probe and a control system; A fixed-point selection module, which uses the front-end transmission unit to transfer the target PCB board to the monitoring area, uses an image acquisition device to collect the surface image of the target PCB board, locates and identifies the plating area to be detected, and adaptively locates the copper thickness detection points; A copper thickness measurement module, which uses the copper thickness detection probe to move to each copper thickness detection point to perform multiple copper thickness detections to obtain a number of collected copper thicknesses, uses an attitude sensor to obtain the tilt angle and direction of the copper thickness detection probe, performs data cleaning on the tilt angle to obtain an effective tilt angle, and inputs it into a pre-set copper thickness detection correction model and combines it with the collected copper thickness to perform copper thickness cosine correction to obtain the monitored copper thickness of each copper thickness detection point; The specific method for adaptively locating the copper thickness detection points is as follows: Obtain the center points of each plating area to be detected, denoted as the position monitoring points corresponding to the plating areas to be detected. Connect the position monitoring points of each plating area to be detected with the position monitoring points of other plating areas to be detected and obtain the connection length as the monitoring distribution distance between each plating area to be detected and other plating areas to be detected; Compare the monitoring distribution distance with the adaptively distributed distance threshold obtained through pre-analysis. Take the plating area to be detected corresponding to the monitoring distribution distance less than the adaptively distributed distance threshold as a suspected associated area, count the number of suspected associated areas corresponding to each plating area to be detected, and then select the plating area to be detected corresponding to the smallest number of suspected associated areas as the initial monitored plating area, and then perform iterative identification of the monitored plating area to obtain each monitored plating area of the target PCB board; Locate the five copper thickness detection points of each monitored plating area according to the pre-set five-point detection method, and summarize to obtain the copper thickness detection points of the target PCB board; The specific construction method of the copper thickness detection correction model is as follows: Use the copper thickness detection probe to perform a limited number of copper thickness detections on several plating areas of several PCB boards to obtain the corresponding copper thickness. At the same time, obtain the tilt angle of the copper thickness detection probe during each copper thickness detection process. Perform pre-correction on each copper thickness and the tilt angle of the corresponding copper thickness detection probe to obtain the monitored copper thickness of each copper thickness detection, and perform deviation analysis with the corresponding copper thickness to obtain the copper thickness deviation degree; Associate the tilt angle of the copper thickness detection probe with the copper thickness deviation degree, and then calculate the average value of the copper thickness deviation degrees corresponding to the same tilt angle of the copper thickness detection probe to obtain the copper thickness deviation degree mapped by the tilt angle of each copper thickness detection probe. Then summarize to obtain the mapping relationship between the tilt angle of the copper thickness detection probe - copper thickness deviation degree, and use it as the copper thickness detection correction model; A defect identification module, which judges whether the target PCB board is abnormal based on the monitored copper thickness, and identifies and outputs the abnormal area.

2. The fully automatic copper thickness measuring instrument for PCB boards based on artificial intelligence according to claim 1, characterized in that: The specific method for identifying the plating area to be detected is as follows: Obtain the chromaticity values of each pixel point on the surface image of the target PCB board, and then compare them with the pre-set chromaticity values of the copper plating to match the set of pixel points with consistent chromaticity values. Then record the continuous pixel points as the same plating area; Perform a coincidence test on each plating layer area with a pre-set target plating layer shape to obtain the area of the coincidence region, and then conduct a matching analysis to obtain the similarity between each plating layer area and the target plating layer shape. Compare with the pre-set similarity threshold to determine whether each plating layer area is a plating layer area to be detected; Record the plating layer areas with similarity greater than the pre-set similarity threshold as the plating layer areas to be detected.

3. The fully automatic copper thickness measuring instrument for PCB boards based on artificial intelligence according to claim 1, characterized in that: The specific method for the adaptive distribution spacing threshold is as follows: Obtain the distances between the position monitoring points of each plating layer area to be detected and the position monitoring points of other plating layer areas to be detected. Then, compare and select the minimum distance as the minimum spacing distance for each plating layer area to be detected, and then calculate the average value to obtain the adaptive distribution spacing threshold.

4. The fully automatic copper thickness measuring instrument for PCB boards based on artificial intelligence according to claim 1, wherein: The specific method for the effective tilt angle is as follows: Conduct a differential comparison on the tilt angles and directions of the copper thickness detection probes during each copper thickness detection process to obtain the tilt angle difference index and direction deviation index between each copper thickness detection and other copper thickness detection processes; Compare the tilt angle difference index and direction deviation index with the pre-set tilt angle difference index threshold and direction deviation index threshold respectively to determine whether the results of each copper thickness detection are valid; Record the tilt angle obtained from the copper thickness detection with a valid judgment result as the effective tilt angle.

5. The fully automatic copper thickness measuring instrument for PCB boards based on artificial intelligence according to claim 1, characterized in that: The specific method for the copper thickness cosine correction is as follows: Input the effective tilt angle into the copper thickness detection correction model to obtain the corresponding copper thickness deviation degree; Multiply the copper thickness deviation degree by the collected copper thickness to obtain the copper thickness correction amount, and then sum it with the collected copper thickness to obtain the monitored copper thickness corresponding to each copper thickness detection point.

6. The fully automatic copper thickness measuring instrument for PCB boards based on artificial intelligence according to claim 5, characterized in that: The specific analysis process for determining whether the target PCB board is abnormal is as follows: Conduct a deviation analysis on the monitored copper thickness of each copper thickness detection point with the pre-set standard copper thickness to obtain the monitored copper thickness deviation degree of each copper thickness detection point, and then compare it with the pre-set monitored copper thickness deviation degree threshold to identify whether there is a copper thickness abnormality at each copper thickness detection point; Count the number of copper thickness detection points identified as having copper thickness abnormalities, and then calculate the proportion of the number of abnormal copper thickness detection points to the total number of copper thickness detection points to obtain the proportion of the number of abnormal copper thickness detection points; Compare the proportion of the number of abnormal copper thickness detection points with the pre-set proportion threshold of the number of abnormal copper thickness detection points to determine whether the target PCB board is abnormal.

7. The fully automatic copper thickness measuring instrument for PCB boards based on artificial intelligence according to claim 6, wherein: [[ID= ​ ​ 8. The fully automatic copper thickness measuring instrument for PCB based on artificial intelligence according to claim 1, characterized in that: ​

Images