Bidirectional result transmission processing method applied to AI and SMT equipment multi-machine cooperation

Abstract

The invention discloses a multi-machine cooperation result two-way transmission processing method applied to AI and SMT equipment, and the method comprises the steps: obtaining the related data of a PCB, storing the data, transmitting the data to an AI server for re-judgment, returning the re-judged result to BRM software, updating the data in a database, displaying the AI re-judgment disqualification information through a re-judgment station interface, and transmitting the AI re-judgment disqualification information to a server. And an operator carries out manual judgment according to the display information. According to the invention, product result data interaction and product result judgment integration are carried out between multiple production lines and multiple devices in an existing SMT production environment and AI big data, so that the manual re-judgment time and the requirement for the misjudgment rate are reduced; the method can reasonably solve the problems of poor stability, poor timeliness and slow data synchronization between devices during transmission of device result data in traditional production, and jamming caused by insufficient memory of the devices due to the adoption of schemes such as local data file generation, can accelerate the timeliness of the whole production system, and improves the production efficiency. The efficiency and the stability of the production system are improved.

Description

Technical Field

[0001] The present invention relates to the field of surface assembly technology, and in particular to a two-way transmission processing method for multi-machine collaboration results applied to AI and SMT equipment. Background Art

[0002] In the SMT electronics manufacturing industry, SMT equipment is widely used to accurately mount electronic components on printed circuit boards (PCBs). However, during the equipment production process, problems such as component position offset and welding defects are difficult to completely avoid. Therefore, in order to ensure product quality, automatic optical inspection (AOI) equipment is usually used on the workshop production line to detect defective products on the mounted circuit boards, and the inspection results are transmitted to the re-inspection station for the operator to conduct timely re-inspection to ensure the qualified efficiency of the circuit boards. However, during the long production process on site, the operator cannot always observe and operate in time, and misjudgment is prone to occur. Therefore, the stability and timeliness of communication and interaction in the entire process will directly affect the overall production efficiency of the production line. The communication interaction in the prior art specifically includes the following shortcomings:

[0003] Poor stability: Due to the limitations of the factory workshop equipment network and LAN IP configuration issues, data communication between devices cannot be particularly stable and continuous. Once a problem occurs on one side, the other side cannot proceed to the next step.

[0004] Low real-time performance: The traditional way of communication between devices is short connection, and usually no persistent connection is established. The connection must be re-established for each request, which is not ideal for real-time and high-frequency communication scenarios.

[0005] Low efficiency: Traditional data transmission methods require re-establishing a connection for each request, which results in a certain performance efficiency overhead, especially in scenarios where data needs to be frequently exchanged between devices.

[0006] Unreliable transmission: Traditional data transmission protocols are stateless and have no reliability guarantee mechanism. Data transmission relies on additional design of the application layer.

[0007] Difficult synchronization: Due to the need to meet the data transmission and processing between multiple production lines, multiple devices and multiple servers, the previous method would increase the difficulty of maintaining system equipment.

[0008] Transmission delay: In the past, data transmission methods required establishing a connection each time, waiting for the result data to be returned after transmitting data, and then closing the connection. Frequent establishment and disconnection processes will increase delays and affect transmission time.

[0009] Large overhead: Excessive overhead will occur during data transmission, and flow control and congestion control mechanisms cannot be implemented to ensure the stability and speed of data flow. Summary of the invention

[0010] The purpose of the present invention is to provide a two-way transmission processing method for multi-machine collaborative results of AI and SMT equipment to solve the problems in the above-mentioned background technology.

[0011] To achieve the above object, the technical solution adopted by the present invention is:

[0012] A bidirectional transmission processing method for AI and multi-machine collaboration results of SMT equipment includes the following steps:

[0013] Obtain data and pictures of defective components generated by AOI equipment scanning PCB boards in SMT production lines;

[0014] The data and pictures of the defective components are processed by communication transmission and stored in a database of a corresponding device in a corresponding format;

[0015] The BRM software on the re-judgment station collects the data and images detected by the current AOI device from the database, processes the data and images, and transmits them to the AI ​​server. The BRM software adds a TCP / IP connection method, which can automatically monitor whether the current corresponding AOI device is started. If it is started, it automatically establishes a two-way communication connection with the AI ​​server;

[0016] The AI ​​server re-judges the received data and images, and returns the re-judgement result data to the BRM software in real time;

[0017] The BRM software decomposes the received result data, updates the data in the database according to the AI ​​judgment results, and transmits the updated data to the re-judgment station;

[0018] The re-judgment station interface displays the bad data that failed the AI ​​re-judgment, and the operator makes manual judgments based on the information displayed on the interface.

[0019] Furthermore, the data of the defective component includes the barcode number in the PCB board, the panel barcode number, the date, the location coordinates of the defective component, the inventory number, the defective component image storage path, and the equipment detection result.

[0020] Furthermore, when the AOI device is scanning and inspecting the PCB board, it can automatically identify whether the PCB board needs AI re-judgment processing. If only manual processing is required and AI re-judgment processing is not required, the AOI device and the AI ​​server are only connected without data transmission. If AI re-judgment processing is required, the data of the PCB board that needs AI re-judgment processing is collected and packaged, and the packaged data is sent to the connected BRM software in real time through a two-way communication request. The BRM software automatically forwards the real-time received data to the AI ​​server.

[0021] Furthermore, the collected data of the PCB board that needs AI re-judgment processing is packaged into a JSON data format.

[0022] Furthermore, the data transmission process between the BRM software and the AI ​​server includes establishing a reliable connection through a three-way handshake for data transmission. After the data transmission is completed, the connection is disconnected through four handshakes to ensure that both parties no longer need to communicate.

[0023] Furthermore, when the BRM software and the AI ​​server are transmitting data, the sent data is numbered by a serial number to ensure that the data packets are transmitted in order and will not be lost or repeated. If a data packet of either party is lost during the transmission process, TCP will automatically retransmit the data packet to ensure data stability. At the same time, both parties perform conversion processing and inspection on the format and content of the data block.

[0024] Furthermore, during the two-way data transmission process, flow control and congestion control are implemented through the receiving window mechanism to ensure efficient and stable data transmission between the AOI device and the AI ​​server.

[0025] Beneficial effects:

[0026] The present invention aims at the demand for product result data interaction and product result judgment integration between multiple production lines and multiple devices in the existing SMT production environment and AI big data to reduce manual re-judgment time and misjudgment rate. The present invention can reasonably solve the problems of poor stability and timeliness when transmitting device result data in traditional production, slow data synchronization between devices, and insufficient memory and freezing of devices caused by using local data file generation solutions. It can speed up the timeliness of the entire production system and improve the efficiency and stability of the production system. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is a flow chart of the present invention.

[0028] Figure 2 It is a data transmission flow chart of the present invention. DETAILED DESCRIPTION

[0029] In order to make the technical problems solved by the present invention, the technical solutions adopted and the technical effects achieved clearer, the technical solutions of the embodiments of the present invention will be further described in detail below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0030] Example

[0031] A bidirectional transmission processing method for AI and multi-machine collaboration results of SMT equipment, such as Figure 1 As shown, the following steps are included:

[0032] S1, obtain the data and pictures of defective components generated by the AOI equipment scanning the PCB board under the SMT production line;

[0033] S2, performing communication transmission processing on the data and pictures of the defective components and storing them in a database of a corresponding device in a corresponding format;

[0034] S3, using the BRM software on the re-judgment station to collect the data and images detected by the current AOI device from the database, and transmit the data and images to the AI ​​server after processing. The BRM software adds a TCP / IP connection mode, which can automatically monitor whether the current corresponding AOI device is started. If it is started, it will automatically establish a two-way communication connection with the AI ​​server;

[0035] S4, the AI ​​server re-judges the received data and images, and returns the re-judgement result data to the BRM software in real time;

[0036] S5, the BRM software decomposes the received result data, updates the data in the database according to the AI ​​judgment results, and transmits the updated data to the re-judgment station;

[0037] S6, the interface of the re-judgment station displays the bad data that failed the AI ​​re-judgment, and the operator makes a manual judgment based on the information displayed on the interface.

[0038] Specifically, every day, the SMT production line is equipped with corresponding AOI equipment to perform automatic optical inspection on each PCB board, detect common defects encountered in welding production, and can collect data and pictures of the corresponding PCB board in real time. The data of the defective component includes the barcode number in the PCB board, the barcode number of the panel, the date, the location coordinates of the defective component, the inventory number, the path to save the picture of the defective component, and the equipment detection results (such as GOOD, NG). After the data detected by these devices are processed, the data communication transmission processing will be carried out in real time and synchronously saved in the local database of the corresponding device. Specifically, when the AOI device collects PCB board data, it adds the TCP / IP connection method to the corresponding supporting software BRM. After adding this connection method, the AOI device can automatically monitor whether the current corresponding device is started. If it is started, it will automatically connect to the remote service of AI re-judgment for two-way communication. During the operation of the device, the two-way communication will always remain connected. Even if the connection is disconnected, the connection request will be automatically re-initiated.

[0039] Among them, when the AOI equipment performs automatic optical inspection on the PCB board, the equipment will store the corresponding data results in the local equipment database by default. The method will also intelligently screen the PCB board information with NG components and store the results in the file location that the workshop operator wants to store, which will be used for the subsequent operator to count and manage the production data. By querying the data of the corresponding equipment, the operator can query the results and other records of the equipment's PCB board inspection based on the query conditions such as the barcode and date of the PCB board, and can also draw charts for the required data through data statistics. By creating indexes for data columns, especially when processing structured data and complex queries, this structured data storage makes the organization, access and management of data more standardized and efficient.

[0040] Preferably, when the AOI device is scanning and testing the PCB board, it can automatically identify whether the PCB board needs AI re-judgment processing. If only manual processing is required and AI re-judgment processing is not required, the AOI device and the AI ​​server are only connected without data transmission. If AI re-judgment processing is required, the data of the PCB board that needs AI re-judgment processing is collected and packaged, and the packaged data is sent to the connected BRM software in real time through a two-way communication request. The BRM software automatically forwards the real-time received data to the AI ​​server. The collected data of the PCB board that needs AI re-judgment processing is packaged into JSON data format.

[0041] Specifically, according to the production rate of different production lines in different time periods, when the AOI device captures the PCB board data information, this method will automatically identify whether the corresponding PCB board needs to be re-judged by AI. If only manual processing is required, the device will remain connected at this time, but the data will not be transmitted, so that the number of device data requests can be dynamically adjusted to ensure the continuous and stable communication between devices. If the workshop production is at a peak period, the personnel can make corresponding settings, and the system will process the data with priority settings, filter out the processing operations of non-set data, close the transmission of such data, and ensure the transmission rate of the set data. The PCB board data that is currently detected in real time and meets the AI ​​re-judgment requirements is intelligently captured, and the corresponding data is synchronously packaged into a JSON data format for two-way transmission. Each piece of data will contain the barcode number scanned and read by the AOI device at this moment, the timestamp of the current time, the number and name of the defective component, the image storage path of the defective component, the device detection result and other fields. The data block is sent to the connected client in real time through a two-way communication request. The client automatically forwards the real-time received data to the AI ​​server. The data transmission method can achieve real-time data transmission continuously, stably and efficiently.

[0042] Preferably, when the BRM software and the AI ​​server are transmitting data, the sent data is numbered by a serial number to ensure that the data packets are transmitted in order and are not lost or repeated. If a data packet of either party is lost during the transmission process, TCP will automatically retransmit the data packet to ensure data stability. At the same time, both parties perform conversion processing and inspection on the format and content of the data block.

[0043] Specifically, the two-way data transmission process between the BRM software and the AI ​​server includes establishing a reliable connection through a three-way handshake for data transmission. During the data transmission process, the sent data is numbered with a sequence number to prevent loss or duplication. After the data transmission is completed, the connection is disconnected through four waves to ensure that both parties no longer need to communicate.

[0044] Among them, a three-way handshake is used to establish a connection. Once established, both parties can send and receive data at the same time. When it is detected that a PCB board is flowing and data needs to be transmitted, the method will promptly process the data and send it to the AI ​​server. The AI ​​server will then return the results of the AI ​​big data re-judgment processing to the client in real time through a response. If the current PCB board is judged to be defective at this time, the relevant specific defective problems will be displayed on the re-judgment station, and the operator will make a manual judgment at this time, which reduces the operator's repeated judgment operations and ensures the production efficiency of the production line.

[0045] After the data transmission is completed, the connection is disconnected through a four-wave mechanism. Specifically: The first wave (FIN): The client (i.e., the BRM software) sends a FIN (end) message to the AI ​​server, indicating that it has completed data transmission and wants to disconnect. The second wave (ACK): After receiving the FIN, the AI ​​server returns an ACK confirmation message, indicating that it has received the disconnect request. The third wave (FIN): The AI ​​server may still have some data to send. After the data is sent, the AI ​​server will also send a FIN message, indicating that it has also completed the data transmission. The fourth wave (ACK): After the client receives the FIN from the AI ​​server, it sends an ACK message to confirm the disconnection. At this point, the connection is officially closed. This method can ensure that when the detection device does not need to communicate data, the corresponding device connection can be closed in time to avoid wasting resources.

[0046] Preferably, during the bidirectional data transmission process, flow control and congestion control are implemented through the receive window mechanism (Receive Window) to ensure efficient and stable data transmission between the AOI device and the AI ​​server. TCP allows the receiver to inform the sender of the amount of data it can process. This prevents the receiver from losing packets due to insufficient processing power. TCP detects network congestion and dynamically adjusts the data transmission rate to avoid data loss or delay caused by network overload.

[0047] Specifically, when each AOI device detects a PCB board, if a component is defective, the corresponding defective component image will be stored locally. At this time, the ability of each device to process images and data in real time will be monitored in real time. If the equipment of a production line faces a large amount of data and a large amount of image storage, the AI-processed images will be automatically cleared in time to release the storage pressure of the equipment. At the same time, if a production line is in a shutdown period, flow control and congestion control mechanisms will be used to ensure the stability and speed of the data flow, ensuring that all devices can perform data processing, transmission and image storage with AI big data under a balanced load. When the production line needs to restart production, as long as the device starts running, it will automatically re-establish a connection with the client and start interacting with the heartbeat packet to ensure the stability of the connection, thereby ensuring the efficiency of the AI ​​re-judgment of the production line. A TCP connection maintained for a long time can transmit data faster because it avoids the connection overhead before each transmission. Once the connection is established, data can be sent directly with less delay and faster speed.

[0048] The bidirectional transmission processing method for the results of multi-machine collaboration between AI and SMT equipment in this embodiment is highly reliable. Before multiple production line equipment and the server communicate bidirectionally, a three-way handshake will be used to establish a bidirectional connection. It ensures that the connection between the two communicating parties is very reliable. Compared with the conventional connection method, this method has a faster connection speed, is quicker and smarter, and does not require the operator to manually perform the connection action. In addition, when facing high-frequency or large-scale data transmission, it saves the overhead of repeatedly establishing connections and supports parallel data communication between multiple production line equipment and servers. Use protocols similar to QUIC (Quick UDP Internet Connection) to reduce connection overhead and better manage and optimize network traffic. At the same time, the improvement and upgrade of security protocols such as TLS / DTLS can provide better encryption and authentication services to ensure the security of workshop data; the work efficiency is higher, and the long connection protocol is adopted. Once the connection is established, the data can be continuously transmitted in both directions without re-establishing the connection each time. After the connection is established, the subsequent transmission speed is faster, which is particularly suitable for big data or application scenarios that require continuous communication. A two-way transmission communication mechanism is adopted. During two-way communication, after each data packet is sent, the receiver will send an ACK (confirmation) signal to the sender to confirm that the data packet has been received. If the sender does not receive the ACK confirmation from the receiver within the specified time, the sender will consider the data packet lost and retransmit it until the data packet arrives successfully. This mechanism ensures the real-time nature of the data, ensures that there will be no data loss, and the receiver can confirm the arrival of each data packet. Reduce the problem of reduced production line efficiency caused by data loss between the equipment and the AI ​​server; it is more convenient for staff to count and query data. During the communication process, based on the relational model, the data will be intelligently stored in a table composed of rows and columns, and the data will be connected through the primary key, foreign key and other relationships of the table. This structured data storage method makes it more standardized and efficient for operators to organize, access and manage the data of PCB boards produced by each production line; it can also adaptively route, automatically adjust the routing and forwarding strategies by learning the behavior and characteristics of the network, and make the protocol communication more intelligent. When a route or port is occupied, it will automatically identify and switch to another port that can be successfully connected.

[0049] The technical principle of the present invention is described above in conjunction with specific embodiments. These descriptions are only for explaining the principle of the present invention and cannot be interpreted as limiting the scope of protection of the present invention in any way. Based on the explanations herein, those skilled in the art can associate other specific implementations of the present invention without paying creative labor, and these methods will fall within the scope of protection of the present invention.

Claims

1. A two-way transmission processing method for multi-machine collaboration results between AI and SMT equipment, characterized in that: The following steps are involved: Obtain data and pictures of defective components generated by AOI equipment scanning PCB boards in SMT production lines; The data and pictures of the defective components are processed by communication transmission and stored in a database of a corresponding device in a corresponding format; The BRM software on the re-judgment station collects the data and images detected by the current AOI device from the database, processes the data and images, and transmits them to the AI ​​server. The BRM software adds a TCP / IP connection method, which can automatically monitor whether the current corresponding AOI device is started. If it is started, it automatically establishes a two-way communication connection with the AI ​​server; The AI ​​server re-judges the received data and images, and returns the re-judgement result data to the BRM software in real time; The BRM software decomposes the received result data, updates the data in the database according to the AI ​​judgment results, and transmits the updated data to the re-judgment station; The re-judgment station interface displays the bad data that failed the AI ​​re-judgment, and the operator makes manual judgments based on the information displayed on the interface.

2. The method for bidirectional transmission of multi-machine collaborative results applied to AI and SMT equipment according to claim 1 is characterized in that: The data of the defective component includes the barcode number in the PCB board, the barcode number of the panel, the date, the location coordinates of the defective component, the inventory number, the path for saving the defective component image, and the equipment detection result.

3. The method for bidirectional transmission of multi-machine collaborative results applied to AI and SMT equipment according to claim 1 is characterized in that: When the AOI device is scanning and inspecting the PCB board, it can automatically identify whether the PCB board needs AI re-judgment processing. If only manual processing is required and AI re-judgment processing is not required, the AOI device and the AI ​​server are only connected without data transmission. If AI re-judgment processing is required, the data of the PCB board that needs AI re-judgment processing is collected and packaged, and the packaged data is sent to the connected BRM software in real time through a two-way communication request. The BRM software automatically forwards the real-time received data to the AI ​​server.

4. The method for bidirectional transmission of multi-machine collaborative results applied to AI and SMT equipment according to claim 3 is characterized in that: The collected data of the PCB board that needs AI re-judgment processing is packaged into JSON data format.

5. The method for bidirectional transmission of multi-machine collaborative results applied to AI and SMT equipment according to claim 1 is characterized in that: The data transmission process between the BRM software and the AI ​​server includes establishing a reliable connection through a three-way handshake for data transmission. After the data transmission is completed, the connection is disconnected through four handshakes to ensure that both parties no longer need to communicate.

6. The method for bidirectional transmission of multi-machine collaborative results applied to AI and SMT equipment according to claim 1 is characterized in that: When the BRM software and the AI ​​server are transmitting data, the sent data is numbered by a serial number to ensure that the data packets are transmitted in order and will not be lost or repeated. If a data packet from either party is lost during the transmission process, TCP will automatically retransmit the data packet to ensure data stability. At the same time, both parties perform conversion processing and inspection on the format and content of the data block.

7. The method for bidirectional transmission of multi-machine collaboration results applied to AI and SMT equipment according to claim 1 is characterized in that: During the two-way data transmission process, flow control and congestion control are implemented through the receiving window mechanism to ensure efficient and stable data transmission between the AOI device and the AI ​​server.

Images