Remote intelligent monitoring method for production workshop
Through remote intelligent monitoring methods, images and temperature data during PCB welding are collected and analyzed, welding deviations are evaluated, and the problem of insufficient monitoring comprehensiveness in the existing technology is solved, and the monitoring efficiency and accuracy of welding quality is improved.
Patent Information
- Application Number
- CN202411488690.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-24
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2044-10-24
AI Technical Summary
In the prior art, the quality monitoring method of PCB printed circuit boards has problems such as low monitoring comprehensiveness, which leads to poor detection accuracy and low detection efficiency.
Provide a remote intelligent monitoring method for production workshops. By continuously collecting the image and temperature distribution data after the PCB is automatically positioned during the welding process, identifying and remotely transmitting positioning deviation and temperature welding deviation information, combining the image information after welding. , conduct positioning and welding deviation fault trend analysis, evaluate welding quality impact level, and realize real-time monitoring and quality evaluation.
It improves the monitoring efficiency and accuracy of welding quality, realizes comprehensive real-time monitoring and quality impact assessment of PCB welding process, and solves the problems of low detection accuracy and efficiency.
Smart Images

Figure CN119057303A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field related to printed circuits, and in particular to a remote intelligent monitoring method for a production workshop. Background Art
[0002] In modern manufacturing, the quality of PCB printed circuit boards is crucial to the performance and reliability of electronic products. In the production process of PCB, the welding quality directly affects the stability and functional performance of the product. Therefore, real-time monitoring and quality control of the PCB welding process become particularly important. Traditional welding quality monitoring methods rely on manual inspection, which has problems such as low inspection efficiency, insufficient accuracy, and slow response speed, and cannot meet the needs of efficient and precise modern production.
[0003] Therefore, in the prior art, the quality monitoring method of PCB printed circuit boards has the technical problems of low comprehensiveness of monitoring, poor detection accuracy, and low detection efficiency. Summary of the invention
[0004] The present application provides a remote intelligent monitoring method for a production workshop, which solves the technical problems that the quality monitoring method of PCB printed circuit boards in the prior art has low comprehensiveness of monitoring, resulting in poor detection accuracy and low detection efficiency. It achieves real-time monitoring of the PCB welding process, and comprehensively analyzes and evaluates the influence of positioning and welding process on product quality, thereby improving the monitoring efficiency and accuracy of welding quality.
[0005] The present application provides a remote intelligent monitoring method for a production workshop, the method comprising: in a PCB welding production workshop, continuously collecting images after the PCB is automatically positioned, obtaining a positioning image sequence, identifying the positioning image, obtaining a positioning deviation information sequence and transmitting it remotely. During the welding process, based on a temperature sensor, monitoring and collecting temperature distribution data in welding, obtaining a temperature distribution sequence, identifying the temperature distribution, obtaining a temperature welding deviation information sequence and transmitting it remotely. After welding is completed, collecting images of the completed welding, obtaining a welding image sequence, identifying the welding image, obtaining an image welding deviation information sequence and a welding quality information sequence and transmitting them remotely. According to the positioning deviation information sequence, performing a positioning deviation fault trend analysis, obtaining positioning deviation fault trend information. According to the temperature welding deviation information sequence and the image welding deviation information sequence, combined with the positioning deviation information sequence, performing a welding deviation fault trend analysis, obtaining welding deviation fault trend information. According to the welding quality information sequence, combined with the positioning deviation fault trend information and the welding deviation fault trend information, performing a differentiation analysis of positioning quality influence and welding quality influence, obtaining a positioning quality influence level and a welding quality influence level, as a remote monitoring result, performing management and control.
[0006] In the implementation method, in the welding production workshop of PCB, images of PCB after automatic positioning are continuously collected to obtain a positioning image sequence, and the positioning images are identified, including: in the welding production workshop of PCB, after the PCB is automatically positioned, the automatically positioned images are collected in sequence to obtain a positioning image sequence. A set of sample positioning images is collected, and the deviation of the PCB from the standard positioning area in each positioning image is marked to obtain a set of sample positioning deviation information. The sample positioning image set and the sample positioning deviation information set are used to train a positioning deviation identifier. The positioning deviation identifier is used to identify multiple positioning images in the positioning image sequence, obtain a positioning deviation information sequence and transmit it remotely.
[0007] In the implementation, during the welding process, based on the temperature sensor, the temperature distribution data in the welding is monitored and collected, a temperature distribution sequence is obtained, and the temperature distribution is identified, including: during the welding process, based on the temperature sensor, the temperature distribution data of the PCB in the welding is monitored and collected in sequence to obtain a temperature distribution sequence. A sample temperature distribution set is collected, and the sample temperature distribution points are marked according to the deviation of the high temperature points from the standard welding temperature in the sample temperature distribution to obtain a sample temperature welding deviation information set. The sample temperature distribution set and the sample temperature welding deviation information set are used to train a temperature welding deviation analyzer. The temperature welding deviation analyzer is used to identify welding deviations of multiple temperature distributions in the temperature distribution sequence, and a temperature welding deviation information sequence is obtained and remotely transmitted.
[0008] In the implementation, after welding is completed, images of the completed welding are collected to obtain a welding image sequence, and the welding images are recognized, including: after welding is completed, images of the completed PCB are collected in sequence to obtain a welding image sequence. A set of sample welding images is collected, and the welding deviation information set of the sample images is obtained according to the deviation amplitude of the welding point in each sample welding image from the standard welding position. The welding deviation information set of the sample image is obtained according to the deviation amplitude between the quality of the welding point and the standard welding point in each sample welding image. The sample welding quality information set is obtained. The sample welding image set and the sample image welding deviation information set are used to train the image welding deviation analysis branch. The sample welding image set and the sample welding quality information set are used to train the welding quality analysis branch, and the welding image recognizer is obtained in combination with the image welding deviation analysis branch. The welding image recognizer is used to recognize multiple welding images in the welding image sequence, and the image welding deviation information sequence and the welding quality information sequence are obtained and transmitted remotely.
[0009] In the implementation method, according to the positioning deviation information sequence, a positioning deviation fault trend analysis is performed to obtain positioning deviation fault trend information, including: according to the automatic positioning monitoring data of the PCB, a sample positioning deviation information sequence set is collected, and according to the change amplitude of the positioning deviation information of each sample positioning deviation information sequence, a sample positioning deviation fault trend information set is set to obtain. The sample positioning deviation information sequence set and the sample positioning deviation fault trend information set are used to construct a positioning deviation trend classifier. The positioning deviation information sequence is input into the positioning deviation trend classifier for classification to obtain positioning deviation fault trend information.
[0010] In the implementation, according to the temperature welding deviation information sequence and the image welding deviation information sequence, combined with the positioning deviation information sequence, welding deviation fault trend analysis is performed to obtain welding deviation fault trend information, including: according to the welding position monitoring data of the PCB, a sample temperature welding deviation information sequence set and a sample image welding deviation information sequence set with positioning deviation information of 0 are collected, and according to the change amplitude of the welding deviation in each sample temperature welding deviation information sequence and the sample image welding deviation information sequence, a sample welding deviation fault trend information set is set. The sample temperature welding deviation information sequence set and the sample image welding deviation information sequence set are respectively used, combined with the sample welding deviation fault trend information set, to construct a temperature welding deviation trend classifier and an image welding deviation trend classifier. The deviation of the positioning deviation information sequence is deleted in the temperature welding deviation information sequence and the image welding deviation information sequence to obtain a separate temperature welding deviation information sequence and a separate image welding deviation information sequence. The change amplitude of the separate temperature welding deviation information sequence and the separate image welding deviation information sequence is respectively input into the temperature welding deviation trend classifier and the image welding deviation trend classifier to obtain the temperature welding deviation fault trend information and the image welding deviation fault trend information. The temperature welding deviation fault trend information and the image welding deviation fault trend information are weightedly calculated to obtain the welding deviation fault trend information.
[0011] In the implementation, according to the welding quality information sequence, combined with the positioning deviation fault trend information and the welding deviation fault trend information, a differentiation analysis of the positioning quality impact and the welding quality impact is performed to obtain the positioning quality impact level and the welding quality impact level as the remote monitoring result, including: according to the welding quality information sequence, analyzing and obtaining the welding quality change trend information. Obtaining the transmission delay of the current remote transmission, combining the welding quality change trend information, and classifying and obtaining the welding quality fault level. According to the size of the positioning deviation fault trend information and the welding deviation fault trend information, the welding quality fault level is differentiated and allocated to obtain the positioning quality impact level and the welding quality impact level, wherein the sum of the positioning quality impact level and the welding quality impact level is the welding quality fault level, and the ratio of the positioning quality impact level to the welding quality impact level is the same as the ratio of the positioning deviation fault trend information to the welding deviation fault trend information. The positioning quality impact level and the welding quality impact level are used as the remote monitoring result.
[0012] In the implementation method, the transmission delay of the current remote transmission is obtained, and the welding quality change trend information is combined to classify and obtain the welding quality fault level, including: collecting a sample transmission delay set, and collecting a sample welding quality change trend information set, and setting a sample welding quality fault level set according to the size of each sample transmission delay and the sample welding quality change trend information, wherein the size of the welding quality fault level and the transmission delay are positively correlated with the size of the welding quality fault level. The sample transmission delay set and the sample welding quality change trend information set are used as decision inputs, and the sample welding quality fault level set is used as decision output. Based on the decision tree, a welding quality classifier is constructed. The welding quality classifier is used to classify the transmission delay and welding quality change trend information input to obtain the welding quality fault level.
[0013] A remote intelligent monitoring method for a production workshop proposed in this application is proposed. In the welding production workshop of the PCB, the image after the PCB is automatically positioned is continuously collected to obtain a positioning image sequence, and the positioning image is identified to obtain a positioning deviation information sequence and transmit it remotely. During the welding process, based on the temperature sensor, the temperature distribution data in the welding is monitored and collected to obtain a temperature distribution sequence, and the temperature distribution is identified to obtain a temperature welding deviation information sequence and transmit it remotely. After the welding is completed, the image of the completed welding is collected to obtain a welding image sequence, the welding image is identified, and the image welding deviation information sequence and the welding quality information sequence are obtained and transmitted remotely. According to the positioning deviation information sequence, the positioning deviation fault trend analysis is performed to obtain the positioning deviation fault trend information. According to the temperature welding deviation information sequence and the image welding deviation information sequence, combined with the positioning deviation information sequence, the welding deviation fault trend analysis is performed to obtain the welding deviation fault trend information. According to the welding quality information sequence, combined with the positioning deviation fault trend information and the welding deviation fault trend information, the differentiation analysis of the positioning quality impact and the welding quality impact is performed to obtain the positioning quality impact level and the welding quality impact level, which are used as the remote monitoring result for management and control. The present invention solves the technical problems that the quality monitoring method of PCB printed circuit boards in the prior art has low comprehensiveness of monitoring, resulting in poor detection accuracy and low detection efficiency. The present invention achieves real-time monitoring of the PCB welding process, and comprehensively analyzes and evaluates the influence of positioning and welding process on product quality, thereby improving the monitoring efficiency and accuracy of welding quality. BRIEF DESCRIPTION OF THE DRAWINGS
[0014] In order to more clearly illustrate the technical solution of the embodiment of the present disclosure, the accompanying drawings of the embodiment of the present disclosure will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the system according to the embodiment of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.
[0015] Figure 1 A flowchart of a remote intelligent monitoring method for a production workshop provided in an embodiment of the present application; Figure 2 A schematic diagram of a flow chart of a remote intelligent monitoring method for a production workshop for obtaining a positioning deviation information sequence provided in an embodiment of the present application. DETAILED DESCRIPTION
[0016] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.
[0017] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.
[0018] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.
[0019] The present application embodiment provides a remote intelligent monitoring method for a production workshop, such as Figure 1 As shown, the method includes: In the PCB welding production workshop, the images of the PCB after automatic positioning are continuously collected to obtain the positioning image sequence, and the positioning image is identified to obtain the positioning deviation information sequence and transmit it remotely. During the welding process, based on the temperature sensor, the temperature distribution data in the welding is monitored and collected to obtain the temperature distribution sequence, and the temperature distribution is identified to obtain the temperature welding deviation information sequence and transmit it remotely. After the welding is completed, the images of the completed welding are collected to obtain the welding image sequence, the welding images are identified, the image welding deviation information sequence and the welding quality information sequence are obtained and transmitted remotely.
[0020] In a PCB welding production workshop, in order to ensure the accuracy of welding, it is necessary to automatically position the PCB, and use a camera to collect images of the PCB after the automatic positioning of the production workshop is completed, and obtain an image sequence of PCB positioning. Based on the positioning image sequence, a positioning deviation information sequence is obtained and transmitted remotely. Subsequently, during the welding process, based on a temperature sensor, the temperature distribution data during welding is monitored and collected in real time to obtain a temperature distribution sequence during the welding process, and the temperature distribution is identified to obtain a temperature welding deviation information sequence and transmit it remotely. Further, after the welding is completed, an image of the completed welding is collected to obtain a welding image sequence, and the welding image contains images of each solder joint of the PCB. The welding image is identified, and an image welding deviation information sequence and a welding quality information sequence are obtained and transmitted remotely, wherein the image welding deviation information is the deviation of the position, and the welding quality information is an evaluation parameter of the solder joint quality.
[0021] like Figure 2 As shown, the method provided by the embodiment of the present application also includes: in a welding production workshop of a PCB, after the PCB is automatically positioned, automatically positioned images are collected in sequence to obtain a positioning image sequence. A set of sample positioning images is collected, and the magnitude of the PCB deviation from the standard positioning area in each positioning image is marked to obtain a sample positioning deviation information set. The sample positioning image set and the sample positioning deviation information set are used to train a positioning deviation identifier. The positioning deviation identifier is used to identify multiple positioning images in the positioning image sequence, obtain a positioning deviation information sequence, and transmit it remotely.
[0022] In a PCB welding production workshop, after the PCB completes automatic positioning, the automatically positioned images are collected in sequence to obtain a positioning image sequence. A sample positioning image set is collected, and the sample positioning image set is the automatically positioned image collected during the historical positioning process. The amplitude of the PCB deviation from the standard positioning area in each positioning image is marked to obtain a sample positioning deviation information set, that is, the amplitude of the PCB deviation from the standard positioning area in each positioning image in the sample positioning image set is marked, and the standard positioning area is fixed with the reference target positioning position to obtain the sample positioning deviation information set. The sample positioning deviation information set contains the deviation parameters of each sample image and the standard positioning area in different directions. When constructing a positioning deviation identifier, based on the sample positioning image set and the sample positioning deviation information set, a convolutional neural network CNN is used to construct a positioning deviation identifier. CNN is good at processing image data and can effectively identify patterns and features in images. The sample positioning image set and the sample positioning deviation information set are input into the convolutional neural network to supervise the model training until the output accuracy of the model meets the predetermined requirements, and the trained model is obtained to obtain a positioning deviation identifier. Finally, the positioning deviation identifier is used to identify multiple positioning images in the positioning image sequence, and a positioning deviation information sequence output by the positioning deviation identifier is obtained, and the positioning deviation information sequence is obtained and remotely transmitted.
[0023] The method provided in the embodiment of the present application also includes: in the welding process, based on the temperature sensor, sequentially monitoring and collecting the temperature distribution data of the PCB during welding, and obtaining a temperature distribution sequence. Collecting a sample temperature distribution set, marking according to the amplitude of the high temperature point in the sample temperature distribution deviating from the standard welding temperature, and obtaining a sample temperature welding deviation information set. Using the sample temperature distribution set and the sample temperature welding deviation information set, a temperature welding deviation analyzer is trained. Using the temperature welding deviation analyzer, welding deviations are identified for multiple temperature distributions in the temperature distribution sequence, and a temperature welding deviation information sequence is obtained and remotely transmitted.
[0024] During the welding process, based on a temperature sensor, which may be an infrared thermal imager or a temperature sensor such as a thermocouple, the temperature distribution data of the PCB during the welding process is monitored and collected in sequence to obtain a temperature distribution sequence, in which the temperature data at each position of the PCB during the welding process is recorded. A sample temperature distribution set is collected, and the sample temperature distribution set is the temperature distribution data collected during the historical welding process. Subsequently, the sample temperature distribution is marked according to the deviation of the high temperature point from the standard welding temperature, that is, the deviation value of the high temperature point from the standard welding temperature is marked to obtain a sample temperature welding deviation information set. The sample temperature distribution set and the sample temperature welding deviation information set are used as training data to supervise the neural network model until the output of the model meets the accuracy threshold, and the training is completed to obtain a temperature welding deviation analyzer. The temperature welding deviation analyzer is used to identify welding deviations of multiple temperature distributions in the temperature distribution sequence, obtain a temperature welding deviation information sequence and transmit it remotely.
[0025] The method provided in the embodiment of the present application also includes: after welding is completed, sequentially collecting images of the welded PCB to obtain a welding image sequence. Collecting a set of sample welding images, marking them according to the amplitude of the deviation of the solder joints in each sample welding image from the standard welding position, and obtaining a set of sample image welding deviation information. Marking them according to the deviation amplitude between the quality of the solder joints in each sample welding image and the quality of the standard solder joints, and obtaining a set of sample welding quality information. Using the sample welding image set and the sample image welding deviation information set, train an image welding deviation analysis branch. Using the sample welding image set and the sample welding quality information set, train a welding quality analysis branch, and obtain a welding image recognizer in combination with the image welding deviation analysis branch. Using the welding image recognizer, identify multiple welding images in the welding image sequence, obtain an image welding deviation information sequence and a welding quality information sequence, and transmit them remotely.
[0026] After welding is completed, images of the welded PCB are collected in sequence to obtain a welding image sequence. A sample welding image set is collected, and the sample welding image set is a PCB welding image collected during the historical welding process. The sample welding image set is annotated according to the deviation amplitude of the solder joint in each sample welding image from the standard welding position, that is, the deviation parameter of the solder joint in each sample welding image from the standard welding position is annotated to obtain a sample image welding deviation information set. Subsequently, the sample welding image is annotated according to the deviation amplitude of the solder joint quality from the standard solder joint quality, and the sample welding quality information set is obtained by manually comparing the welding degree, solder joint size, shape, etc. of each sample welding image solder joint with the quality deviation parameters corresponding to the image annotation of the standard solder joint quality. The sample welding image set and the sample image welding deviation information set are used to train the image welding deviation analysis branch. The sample welding image set and the sample welding quality information set are used to train the welding quality analysis branch, and the welding image recognizer is obtained by combining the image welding deviation analysis branch. Among them, the construction method of the welding quality analysis branch and the welding deviation analysis branch is consistent with the construction method of the positioning deviation recognizer, and the constructed welding quality analysis branch and the welding deviation analysis branch are obtained, and combined to obtain a welding image recognizer. The welding image recognizer is used to output corresponding welding deviation data and welding quality deviation data according to the welding image. The welding image recognizer is used to recognize multiple welding images in the welding image sequence, obtain image welding deviation information sequence and welding quality information sequence and transmit them remotely.
[0027] The method provided in the embodiment of the present application also includes: performing positioning deviation fault trend analysis according to the positioning deviation information sequence to obtain positioning deviation fault trend information. Performing welding deviation fault trend analysis according to the temperature welding deviation information sequence and the image welding deviation information sequence in combination with the positioning deviation information sequence to obtain welding deviation fault trend information. Performing differentiation analysis of positioning quality influence and welding quality influence according to the welding quality information sequence in combination with the positioning deviation fault trend information and the welding deviation fault trend information to obtain positioning quality influence level and welding quality influence level as remote monitoring results for management and control.
[0028] According to the positioning deviation information sequence, a positioning deviation fault trend analysis is performed to obtain positioning deviation fault trend information, and the positioning deviation fault trend information is the deviation trend information generated by automatic positioning. According to the temperature welding deviation information sequence and the image welding deviation information sequence, combined with the positioning deviation information sequence, a welding deviation fault trend analysis is performed to obtain welding deviation fault trend information, and the welding deviation fault trend information is the welding deviation trend generated by welding temperature and welding deviation. Finally, according to the welding quality information sequence, combined with the positioning deviation fault trend information and the welding deviation fault trend information, a differentiation analysis of positioning quality influence and welding quality influence is performed to obtain the positioning quality influence level and the welding quality influence level. As the remote monitoring result, the optimization guidance of the technician is given according to the size of the positioning quality influence level and the welding quality influence level, so as to facilitate the control of the equipment. If the positioning deviation is evaluated as a high-level influence and the welding deviation is a low-level influence, the positioning problem is solved first. The quality monitoring method of the PCB printed circuit board in the prior art solves the technical problems that the comprehensiveness of monitoring is low, resulting in poor detection accuracy and low detection efficiency. It achieves real-time monitoring of the PCB welding process, and comprehensively analyzes and evaluates the impact of positioning and welding processes on product quality, thereby improving the monitoring efficiency and accuracy of welding quality.
[0029] The method provided in the embodiment of the present application also includes: according to the automatic positioning monitoring data of the PCB, a set of sample positioning deviation information sequences is collected, and according to the change amplitude of the positioning deviation information in each sample positioning deviation information sequence, a set of sample positioning deviation fault trend information is set to obtain. The sample positioning deviation information sequence set and the sample positioning deviation fault trend information set are used to construct a positioning deviation trend classifier. The change amplitude of the positioning deviation information sequence is input into the positioning deviation trend classifier to obtain positioning deviation fault trend information.
[0030] According to the automatic positioning monitoring data of the PCB, the automatic positioning monitoring data of the PCB is the monitoring data of the historical automatic positioning, and a set of sample positioning deviation information sequences is collected. The set of sample positioning deviation information sequences is a set composed of multiple sample positioning deviation information sequences generated by the historical automatic positioning. And according to the change amplitude of the positioning deviation information of each sample positioning deviation information sequence, the change amplitude is the average deviation amplitude between the parameters in each sample positioning deviation information sequence. For example, if the positioning deviation of a certain sample PCB gradually increases from the first to the third amplitude of 2mm, then the change amplitude is 1mm / piece, and the sample positioning deviation fault trend information set is obtained based on the change amplitude setting. Subsequently, the sample positioning deviation information sequence set and the sample positioning deviation fault trend information set are used to construct a positioning deviation trend classifier. The positioning deviation trend classifier is constructed based on a decision tree model, and the sample positioning deviation information sequence set and the sample positioning deviation fault trend information set are used as training data to train the decision tree model, and the trained model is obtained to obtain a positioning deviation trend classifier. The positioning deviation trend classifier is used to output the corresponding positioning deviation fault trend information according to the positioning deviation information sequence. The positioning deviation information sequence is input into the positioning deviation trend classifier for classification to obtain positioning deviation fault trend information.
[0031] The method provided in the embodiment of the application also includes: according to the welding position monitoring data of the PCB, collecting a sample temperature welding deviation information sequence set and a sample image welding deviation information sequence set with positioning deviation information of 0, and setting a sample welding deviation fault trend information set according to the change amplitude of the welding deviation in each sample temperature welding deviation information sequence and sample image welding deviation information sequence. The sample temperature welding deviation information sequence set and the sample image welding deviation information sequence set are respectively used, and the sample welding deviation fault trend information set is combined to construct a temperature welding deviation trend classifier and an image welding deviation trend classifier. The deviation of the positioning deviation information sequence is deleted in the temperature welding deviation information sequence and the image welding deviation information sequence to obtain a separate temperature welding deviation information sequence and a separate image welding deviation information sequence. The change amplitude of the separate temperature welding deviation information sequence and the separate image welding deviation information sequence is respectively input into the temperature welding deviation trend classifier and the image welding deviation trend classifier to obtain the temperature welding deviation fault trend information and the image welding deviation fault trend information. The temperature welding deviation fault trend information and the image welding deviation fault trend information are weighted and calculated to obtain the welding deviation fault trend information.
[0032] According to the welding position monitoring data of the PCB, the welding position monitoring data of the PCB records the welding position of each PCB welding process, and then collects the sample temperature welding deviation information sequence set and the sample image welding deviation information sequence set with the positioning deviation information of 0, and the above sample set is a sample excluding the positioning deviation. And according to the change amplitude of the welding deviation in each sample temperature welding deviation information sequence and the sample image welding deviation information sequence, the sample welding deviation fault trend information set is set to obtain, and the change amplitude of the welding deviation is the average deviation amplitude of the welding deviation sequence. For example, for a certain sample, there are multiple deviation parameters, which are right deviation 1mm, right deviation 2mm, and right deviation 3mm, respectively. At this time, the corresponding average deviation amplitude is right deviation 1mm / piece. According to the change amplitude of the welding deviation, the sample welding deviation fault trend information set is set to obtain. The sample temperature welding deviation information sequence set and the sample image welding deviation information sequence set are respectively used, and the sample welding deviation fault trend information set is combined with the sample welding deviation fault trend information set to construct a temperature welding deviation trend classifier and an image welding deviation trend classifier. The construction method of the temperature welding deviation trend classifier and the image welding deviation trend classifier is consistent with the construction method of the positioning deviation trend classifier, and there is only a difference in training data. The training data of the temperature welding deviation trend classifier is a set of sample temperature welding deviation information sequences and a set of sample welding deviation fault trend information. The training data of the image welding deviation trend classifier is a set of sample image welding deviation information sequences and a set of sample welding deviation fault trend information. The deviation of the positioning deviation information sequence is deleted in the temperature welding deviation information sequence and the image welding deviation information sequence to obtain a separate temperature welding deviation information sequence and a separate image welding deviation information sequence. The elements in the separate temperature welding deviation information sequence are consistent with the elements of the temperature welding deviation information sequence, and the separate image welding deviation information sequence is a sequence after deleting the deviation of the positioning deviation information sequence in the image welding deviation information sequence, that is, the deviation in the image welding deviation information is compensated and corrected by the deviation of the positioning deviation information sequence. For example, if there is a right deviation of 3mm in the original image welding deviation information, and the PCB deviation corresponding to the positioning deviation information is a right deviation of 3mm, then the deviation of the corrected image welding deviation information is 0. The change amplitudes of the separate temperature welding deviation information sequence and the separate image welding deviation information sequence are respectively input into the temperature welding deviation trend classifier and the image welding deviation trend classifier to obtain the temperature welding deviation fault trend information and the image welding deviation fault trend information. Finally, the temperature welding deviation fault trend information and the image welding deviation fault trend information are weightedly calculated, wherein the weights are set based on the actual situation to obtain the welding deviation fault trend information.
[0033] The method provided in the embodiment of the present application also includes: analyzing and obtaining welding quality change trend information according to the welding quality information sequence. Obtaining the transmission delay of the current remote transmission, combining the welding quality change trend information, and classifying and obtaining the welding quality fault level. According to the size of the positioning deviation fault trend information and the welding deviation fault trend information, the welding quality fault level is differentiated and allocated to obtain the positioning quality impact level and the welding quality impact level, wherein the sum of the positioning quality impact level and the welding quality impact level is the welding quality fault level, and the ratio of the positioning quality impact level to the welding quality impact level is the same as the ratio of the positioning deviation fault trend information to the welding deviation fault trend information. The positioning quality impact level and the welding quality impact level are used as remote monitoring results.
[0034] According to the welding quality information sequence, each welding quality information is extracted and the deviation is calculated to obtain the quality deviation of the sequence, and a polynomial fitting is performed on the quality deviation of the sequence, and the welding quality change trend information is obtained based on the fitting result. The transmission delay of the current remote transmission is obtained, and the welding quality fault level is obtained by classification in combination with the welding quality change trend information. Further, the positioning deviation fault trend information and the welding deviation fault trend information are standardized, and the welding quality fault level is differentiated and allocated according to the size of the standardized positioning deviation fault trend information and the welding deviation fault trend information, and the positioning quality impact level and the welding quality impact level are obtained. The positioning quality impact level is the ratio of the standardized positioning deviation fault trend information to the sum of the standardized welding deviation fault trend information and the welding deviation fault trend information, multiplied by the welding quality fault level. The welding quality impact level is the ratio of the standardized positioning deviation fault trend information to the sum of the standardized positioning deviation fault trend information and the welding deviation fault trend information, multiplied by the welding quality fault level. The sum of the positioning quality impact level and the welding quality impact level is the welding quality failure level, and the ratio of the positioning quality impact level to the welding quality impact level is the same as the ratio of the positioning deviation failure trend information to the welding deviation failure trend information. The positioning quality impact level and the welding quality impact level are used as remote monitoring results.
[0035] The method provided in the embodiment of the present application also includes: collecting a set of sample transmission delays, and collecting a set of sample welding quality change trend information, and setting a set of sample welding quality fault levels according to the size of each sample transmission delay and sample welding quality change trend information, wherein the size of the welding quality fault level is positively correlated with the size of the transmission delay and the welding quality fault level. The sample transmission delay set and the sample welding quality change trend information set are used as decision inputs, and the sample welding quality fault level set is used as decision output, and a welding quality classifier is constructed based on a decision tree. The welding quality classifier is used to classify the transmission delay and welding quality change trend information input to obtain the welding quality fault level.
[0036] When obtaining the welding quality fault level, a set of sample transmission delays and a set of sample welding quality change trend information are collected. According to the size of each sample transmission delay and the sample welding quality change trend information, a set of sample welding quality fault levels is obtained through professional and technical personnel. The sample is a data sample of historical transmission. Among them, the size of the welding quality fault level and the transmission delay are positively correlated with the size of the welding quality fault level. Subsequently, the sample transmission delay set and the sample welding quality change trend information set are used as decision inputs, and the sample welding quality fault level set is used as decision output. Based on the decision tree, a welding quality classifier is constructed. When the output of the trained welding classifier meets the preset accuracy rate, the trained welding quality classifier is obtained. Finally, the welding quality classifier is used to input and classify the transmission delay and welding quality change trend information to obtain the welding quality fault level.
[0037] The technical solution provided by the embodiment of the present invention obtains a positioning deviation information sequence by identifying the collected positioning image, collects temperature distribution data based on the temperature sensor, identifies the temperature distribution, and obtains a temperature welding deviation information sequence. A welding image sequence after welding is collected and identified to obtain an image welding deviation information sequence and a welding quality information sequence. According to the positioning deviation information sequence, positioning deviation fault trend information is obtained, and welding deviation fault trend information is obtained by combining the temperature welding and image welding deviation information sequences. According to the welding quality information sequence, the positioning deviation fault trend information and the welding deviation fault trend information are combined to obtain the positioning quality impact level and the welding quality impact level. The technical problem that the quality monitoring method of PCB printed circuit boards in the prior art has low comprehensiveness of monitoring, resulting in poor detection accuracy and low detection efficiency is solved. The PCB welding process is monitored in real time, and the impact level of the positioning and welding processes on the quality of the product is comprehensively analyzed and evaluated, thereby improving the monitoring efficiency and accuracy of the welding quality.
[0038] Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.
Claims
1. A remote intelligent monitoring method for a production workshop, characterized in that: The method comprises: In the PCB welding production workshop, the images of the PCB after automatic positioning are continuously collected to obtain the positioning image sequence, and the positioning image is identified to obtain the positioning deviation information sequence and transmit it remotely; During the welding process, the temperature distribution data in welding is monitored and collected based on the temperature sensor, the temperature distribution sequence is obtained, and the temperature distribution is identified, and the temperature welding deviation information sequence is obtained and transmitted remotely; After welding is completed, the welding image is collected to obtain a welding image sequence, the welding image is identified, and the image welding deviation information sequence and welding quality information sequence are obtained and transmitted remotely; According to the positioning deviation information sequence, performing positioning deviation fault trend analysis to obtain positioning deviation fault trend information; According to the temperature welding deviation information sequence and the image welding deviation information sequence, combined with the positioning deviation information sequence, welding deviation fault trend analysis is performed to obtain welding deviation fault trend information; According to the welding quality information sequence, combined with the positioning deviation fault trend information and the welding deviation fault trend information, a differentiation analysis of the positioning quality impact and the welding quality impact is performed to obtain the positioning quality impact level and the welding quality impact level, which are used as remote monitoring results for management and control.
2. The remote intelligent monitoring method for a production workshop according to claim 1, characterized in that: In the PCB welding production workshop, continuously collect images of the PCB after automatic positioning, obtain the positioning image sequence, and identify the positioning image, including: In a PCB welding production workshop, after the PCB is automatically positioned, automatically positioned images are collected in sequence to obtain a positioning image sequence; Collect a set of sample positioning images, mark the extent to which the PCB in each positioning image deviates from the standard positioning area, and obtain a set of sample positioning deviation information; Using the sample positioning image set and the sample positioning deviation information set to train a positioning deviation identifier; The positioning deviation identifier is used to identify multiple positioning images in the positioning image sequence, obtain a positioning deviation information sequence and transmit it remotely.
3. The remote intelligent monitoring method for a production workshop according to claim 1, characterized in that: During the welding process, the temperature distribution data in the welding process is monitored and collected based on the temperature sensor, the temperature distribution sequence is obtained, and the temperature distribution is identified, including: During the welding process, the temperature distribution data of the PCB during welding is monitored and collected in sequence based on the temperature sensor to obtain the temperature distribution sequence; Collect a sample temperature distribution set, mark the sample temperature distribution according to the deviation of the high temperature point from the standard welding temperature, and obtain the sample temperature welding deviation information set; Using the sample temperature distribution set and the sample temperature welding deviation information set to train a temperature welding deviation analyzer; The temperature welding deviation analyzer is used to identify welding deviations of multiple temperature distributions in the temperature distribution sequence, and a temperature welding deviation information sequence is obtained and remotely transmitted.
4. The remote intelligent monitoring method for a production workshop according to claim 1, characterized in that: After welding is completed, the completed welding images are collected to obtain the welding image sequence, and the welding images are recognized, including: After the welding is completed, the images of the welded PCB are collected in sequence to obtain a welding image sequence; Collecting a set of sample welding images, marking the welding points in each sample welding image according to the deviation from the standard welding position, and obtaining a set of sample image welding deviation information; Marking is performed according to the deviation between the quality of the weld spot in each sample welding image and the quality of the standard weld spot, so as to obtain a sample welding quality information set; Using the sample welding image set and the sample image welding deviation information set to train an image welding deviation analysis branch; The sample welding image set and the sample welding quality information set are used to train a welding quality analysis branch, and combined with the image welding deviation analysis branch to obtain a welding image recognizer; The welding image recognizer is used to recognize multiple welding images in the welding image sequence, and an image welding deviation information sequence and a welding quality information sequence are obtained and transmitted remotely.
5. The remote intelligent monitoring method for a production workshop according to claim 1, characterized in that: According to the positioning deviation information sequence, positioning deviation fault trend analysis is performed to obtain positioning deviation fault trend information, including: According to the automatic positioning monitoring data of the PCB, a sample positioning deviation information sequence set is collected, and according to the change amplitude of the positioning deviation information of each sample positioning deviation information sequence, a sample positioning deviation fault trend information set is set to obtain; Using the sample positioning deviation information sequence set and the sample positioning deviation fault trend information set, constructing a positioning deviation trend classifier; The positioning deviation information sequence is input into the positioning deviation trend classifier for classification to obtain positioning deviation fault trend information.
6. The remote intelligent monitoring method for a production workshop according to claim 1, characterized in that: According to the temperature welding deviation information sequence and the image welding deviation information sequence, combined with the positioning deviation information sequence, welding deviation fault trend analysis is performed to obtain welding deviation fault trend information, including: According to the soldering position monitoring data of the PCB, a sample temperature soldering deviation information sequence set and a sample image soldering deviation information sequence set with positioning deviation information of 0 are collected, and according to the change amplitude of the soldering deviation in each sample temperature soldering deviation information sequence and sample image soldering deviation information sequence, a sample soldering deviation fault trend information set is set to obtain; The sample temperature welding deviation information sequence set and the sample image welding deviation information sequence set are respectively used, and combined with the sample welding deviation fault trend information set, a temperature welding deviation trend classifier and an image welding deviation trend classifier are constructed; Deleting the deviation of the positioning deviation information sequence in the temperature welding deviation information sequence and the image welding deviation information sequence to obtain a separate temperature welding deviation information sequence and a separate image welding deviation information sequence; Inputting the change amplitudes of the separate temperature welding deviation information sequence and the separate image welding deviation information sequence into the temperature welding deviation trend classifier and the image welding deviation trend classifier respectively, to obtain the temperature welding deviation fault trend information and the image welding deviation fault trend information; The temperature welding deviation fault trend information and the image welding deviation fault trend information are weightedly calculated to obtain the welding deviation fault trend information.
7. The remote intelligent monitoring method for a production workshop according to claim 1, characterized in that: According to the welding quality information sequence, combined with the positioning deviation fault trend information and the welding deviation fault trend information, a differentiation analysis of the positioning quality impact and the welding quality impact is performed to obtain the positioning quality impact level and the welding quality impact level as the remote monitoring result, including: Analyze and obtain welding quality change trend information according to the welding quality information sequence; Obtaining the transmission delay of the current remote transmission, combining the welding quality change trend information, and classifying and obtaining the welding quality fault level; According to the sizes of the positioning deviation fault trend information and the welding deviation fault trend information, the welding quality fault level is differentiated and allocated to obtain a positioning quality impact level and a welding quality impact level, wherein the sum of the positioning quality impact level and the welding quality impact level is the welding quality fault level, and the ratio of the positioning quality impact level to the welding quality impact level is the same as the ratio of the positioning deviation fault trend information to the welding deviation fault trend information; The positioning quality impact level and the welding quality impact level are used as remote monitoring results.
8. The remote intelligent monitoring method for a production workshop according to claim 7, characterized in that: The transmission delay of the current remote transmission is obtained, and the welding quality fault level is obtained by classification in combination with the welding quality change trend information, including: Collecting a set of sample transmission delays and a set of sample welding quality change trend information, and setting a set of sample welding quality fault levels according to the size of each sample transmission delay and the sample welding quality change trend information, wherein the size of the welding quality fault level is positively correlated with the size of the transmission delay and the welding quality fault level; Using the sample transmission delay set and the sample welding quality change trend information set as decision inputs, using the sample welding quality fault level set as decision output, and constructing a welding quality classifier based on a decision tree; The welding quality classifier is used to input and classify the transmission delay and welding quality change trend information to obtain the welding quality fault level.
Citation Information
Patent Citations
Intelligent online detection device based on multi-source information fusion for welded joint
CN109483107A
PCB defect identification method applied to PCB automatic sorting process
CN114882039A
Tin soldering process optimization method and system for improving welding precision
CN116423005A
Power control method and system of handheld laser welding machine
CN117260035A
Temperature control method and system of high-speed automatic wire bonder
CN117733415A
Cited By
Method and system for monitoring dynamic burning state of PCB (Printed Circuit Board) card
CN120525882A
