A high pressure die casting die island casting piece traceability control system and method

Abstract

The present application relates to the technical field of industrial automation control, and especially relates to a high-pressure casting die-casting island casting traceability control system, which mainly solves the problems of difficult accurate correlation of heterogeneous equipment data and difficult defect root cause traceability in high-frequency die-casting production. The system comprises: an MES-SystemServo processing module, which acquires a product code and creates an assembly data table containing DBBlock, Address and Flag flag bits for bottom layer logic mapping; a heterogeneous communication protocol module, which dynamically collects production data through an OPC UA and a MODBUS-TCP double protocol stack based on the above fields; a die-casting island management system module, which associates the collected data with the product code time sequence, and accurately traces back the historical process data before and after the abnormal time point when the alarm is triggered; and a system main control picture module, which binds the traced back data with the abnormal code to generate complete traceability records, and updates the LED look board in linkage.

Description

Technical Field

[0001] This invention relates to the field of industrial automation control technology, and in particular to a traceability control system for castings on high-pressure die-casting islands. Background Technology

[0002] High-pressure die casting plays an irreplaceable role in the production of complex and precision metal parts. However, with the development of the Industrial Internet of Things (IIoT), traditional die-casting island control systems have gradually revealed their shortcomings, such as low efficiency in low-level data acquisition and high communication barriers between heterogeneous devices. Existing systems often employ single or closed communication protocols, making it difficult to achieve efficient concurrent exchange and accurate low-level memory mapping between the high-frequency transient process parameters generated in the underlying programmable logic controller (PLC) and the historical management data of the upper-level manufacturing execution system (MAS). This easily leads to data interaction conflicts, packet loss, and timing misalignments under extremely short die-casting production cycles.

[0003] In the quality control stage, most existing systems only operate at the level of simple threshold alarms, lacking dynamic data traceability and deep binding capabilities anchored by product codes. When a part experiences a quality failure, existing systems cannot automatically extract the precise anomaly trigger time and trace back the key process parameters before and after that time window. This results in a severe disconnect between alarm information and the root cause of the defect, and isolated subsystems struggle to achieve low-latency hardware linkage based on alarm signals. In summary, existing technologies have significant technical bottlenecks in heterogeneous protocol data collaborative addressing and acquisition, precise timing correlation of transient process parameters, and cross-interface anomaly backtracking linkage, failing to meet the urgent needs of modern high-pressure casting for precise traceability of high-frequency data. There is an urgent need in the market for a high-pressure casting die-casting island casting traceability control system and method to solve these problems. Summary of the Invention

[0004] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution: a high-pressure die-casting island casting traceability control system, comprising: The MES-SystemServo processing module is configured to acquire the product code of the casting at the beginning of the die casting cycle and create an assembly data table for data acquisition and processing; the assembly data table includes a DBBlock data block field for underlying mapping, an Address field, and a Flag bit to control the read and write status; The heterogeneous communication protocol module integrates OPC UA and MODBUS-TCP protocols; based on the Flag bit, it uses the DBBlock and Address fields to address the PLC, collects key production data of the die-casting process in real time through MODBUS-TCP, and exchanges historical data through OPC UA. The die-casting island management system module includes multiple integrated subsystem interfaces, configured to correlate the collected production data with the product code in a time sequence; and when a fault alarm is triggered during quality control, it extracts the abnormal time point and backtracks the historical process data before and after that time point; The system's main control screen module, serving as a comprehensive entry point for communication and control, is configured to bind retrospective transient process data with abnormal product codes to generate traceability records, and to update the status of the LED dashboard used for scanning based on alarm signals.

[0005] Preferably, the MES-SystemServo processing module obtains the product code of the casting at the beginning of the die-casting cycle, specifically including: After detecting the die casting cycle start signal output by the die casting equipment, the product code of the current casting is read, and a correspondence is established between the product code and the start time of the current die casting cycle to generate a traceability index corresponding to the current casting.

[0006] Preferably, the DBBlock data block field, Address field, and Flag field in the assembly data table are used to establish the PLC data mapping relationship corresponding to the product code, specifically including: The PLC data block is determined based on the DBBlock data block field, the data address in the PLC data block is determined based on the Address field, and the read / write status of the corresponding data item is determined based on the Flag bit, so that the MES-SystemServo processing module can read or write production data according to the data mapping relationship.

[0007] Preferably, the heterogeneous communication protocol module collects key production data of the die-casting process in real time via MODBUS-TCP, specifically including: Obtain the die-casting process status word output by the PLC, and determine the current die-casting stage based on the die-casting process status word; The first acquisition cycle is used during the mold closing and fast injection stages, and the second acquisition cycle is used during the pressure holding, mold opening and standby stages. The first acquisition cycle is shorter than the second acquisition cycle.

[0008] Preferably, the heterogeneous communication protocol module exchanges historical data via OPC UA, specifically including: Historical production data obtained through OPC UA is written to the local sequence cache pool according to the device identifier and collection time; When a communication connection anomaly is detected, the historical production data that has not been transmitted is retained in the local sequence cache pool and transmitted in the order of the collection time after the communication connection is restored.

[0009] Preferably, the die-casting island management system module performs a time-series association between the collected production data and the product code, specifically including: A hardware timestamp is added to the production data, and the data association time interval is determined by the start time and end time of the die-casting cycle corresponding to the product code. Production data whose collection time is within the data association time interval is associated with the corresponding product code.

[0010] Preferably, when the die-casting island management system module triggers a fault alarm during quality control, it backslides the historical process data, specifically including: Obtain the trigger time of the fault alarm and determine the historical data backtracking time window based on the process parameters corresponding to the fault alarm; When the fault alarm corresponds to an abnormal mold temperature, the time window is extended after the fault alarm trigger time; When the fault alarm corresponds to an abnormal injection pressure, the time window is shortened.

[0011] Preferably, the system main control screen module binds the retrospective transient process data with the abnormal product code to generate a traceability record, specifically including: The transient process data, abnormal product code, and corresponding abnormal occurrence time are obtained. A hash digest algorithm is used to perform a digest calculation on the transient process data, abnormal product code, and abnormal occurrence time to generate a corresponding traceability digest value. The traceability digest value is then associated with and stored with the traceability record.

[0012] Preferably, the system's main control screen module updates the status of the LED signboard used for scanning based on alarm signals, specifically including: The fault alarm is converted into corresponding dashboard status information, and the dashboard status information is sent to the LED dashboard interface via an asynchronous message event bus, so that the LED dashboard interface updates the displayed content according to the received dashboard status information.

[0013] A method for traceability control of die-casting island castings in high-pressure die casting includes: S1: At the start of the die casting cycle, obtain the product code of the casting and create an assembly data table for data acquisition and processing; the assembly data table includes a DBBlock data block field for underlying mapping, an Address field, and a Flag bit to control the read and write status; S2: Based on the Flag bit, the PLC is addressed using the DBBlock data block field and the Address field. Key production data of the die-casting process are collected in real time through the integrated MODBUS-TCP protocol, and historical data is exchanged through the integrated OPCUA protocol. S3: The collected production data is sequentially associated with the product code; and when a fault alarm is triggered during quality control, the abnormal time point is extracted and the historical process data before and after that time point is traced back. S4: Bind the traced transient process data with the abnormal product code to generate a traceability record, and update the status of the LED Kanban used for scanning based on the alarm signal.

[0014] Compared with the prior art, the present invention has the following beneficial effects: This invention achieves anti-collision high-frequency data acquisition and precise low-level addressing in heterogeneous industrial control environments. Addressing the shortcomings of existing technologies, such as high communication barriers and susceptibility to data interaction conflicts between heterogeneous devices, this system integrates OPC UA and MODBUS-TCP dual-protocol stacks by creating an assembly data table containing data block fields, address fields, and read / write flags. This mechanism enables precise mapping and protocol routing of the underlying PLC based on flag status, ensuring efficient concurrent interaction between high-frequency transient process parameters of the die-casting machine and upper-level historical management data, and avoiding data overwriting, packet loss, and timing misalignment issues under extremely short die-casting production cycles.

[0015] This invention achieves precise root cause tracing of defects based on time windows and deep binding of multi-dimensional data. Addressing the bottleneck of existing systems that only provide shallow threshold alarms and cannot pinpoint the root cause of defects, this invention introduces a dynamic data backtracking mechanism based on product codes. When the system triggers a quality fault alarm, it can automatically extract the precise time point of the anomaly trigger, backtrack to the key historical process data within that time window, and irreversibly bind the actual equipment status, such as transient pressure and temperature, to the specific product code where the anomaly occurred.

[0016] This invention achieves low-latency state linkage and hardware-level isolated response across subsystem interfaces. Addressing the shortcomings of existing die-casting island systems where modules are isolated and difficult to implement abnormal linkage responses, this system constructs a multi-integrated subsystem interface architecture. When the underlying layer performs quality control and issues a fault alarm, the system not only generates a traceability record but also directly triggers the LED dashboard status update for scanning based on the alarm signal across interfaces. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the traceability control system for high-pressure die casting islands according to the present invention; Figure 2 This is a schematic diagram of a traceability control method for high-pressure die casting islands according to the present invention; Figure 3 This is a diagram illustrating the underlying data mapping and read / write interaction logic of this invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Example

[0019] Please see Figures 1 to 3 As shown, this embodiment provides a traceability control system for die-casting islands in high-pressure die casting, which specifically includes the following steps: In this embodiment, the product code of the casting is obtained at the beginning of the die-casting cycle. The specific implementation method is as follows: The die-casting cycle start signal output by the die-casting equipment serves as the physical trigger for the entire traceability process. The MES-SystemServo processing module continuously monitors the specific address bit corresponding to the die-casting machine's mold-closing action by establishing a high-frequency heartbeat polling mechanism with the die-casting island's programmable logic controller. To avoid false triggering caused by electromagnetic interference or relay mechanical jitter in the industrial environment, the system introduces a preset signal anti-jitter time threshold after detecting a low-to-high level transition at this address bit. This signal anti-jitter time threshold refers to the shortest time required for the system to continuously monitor the level state after initially detecting a level transition and ensure it remains unchanged. Only when the duration of the high-level state completely covers and exceeds this anti-jitter time threshold does the system confirm that the die-casting cycle start signal is genuine and valid.

[0020] Upon confirmation of the start signal, the MES-SystemServo processing module immediately sends a read command to the industrial vision recognition terminal or RFID antenna deployed at the mold opening station. The reading device acquires the physical tag information on the casting carrier currently inside the mold cavity or being grasped by the robotic arm, decodes it into a standard format product code string, and uploads it to the server.

[0021] The system establishes a correspondence between the product code and the start time of the current die-casting cycle. In this step, to ensure absolute alignment of cross-device data on the timeline, the system server and the underlying hardware control network pre-synchronize with a high-precision clock using a precise time protocol, keeping the time deviation between devices within the microsecond range. Upon confirming the validity of the start signal, the system directly extracts the absolute timestamp from the underlying system as the start time of the current die-casting cycle, and pushes this start time and the read product code as two parallel data elements into a newly created data row in memory.

[0022] Generating the traceability index for the current casting is the final step in this process. Following preset coding rules, the system concatenates a fixed prefix representing the die-casting island number, a formatted start-time string, and the serial number portion of the product code to generate a globally unique string identifier throughout the entire manufacturing execution system. This traceability index will serve as the core primary key for binding all subsequent process parameters, quality inspection results, and anomaly alarm data, accompanying the casting throughout its entire lifecycle.

[0023] Taking the die-casting production line of aluminum alloy cylinder blocks for automobile engines as an example, the programmable logic controller inside the die-casting island sets the 20th byte of the DB10 data block as the signal mapping address for the completion of the mold closing action. When the die-casting machine's mold closing mechanism reaches its position, the value of this address changes from 0 to 1. The MES-SystemServo processing module detects this change at a polling cycle of 10 milliseconds and starts a timer. The system's preset signal debouncing time threshold is 50 milliseconds. If the value of this address remains stable at 1 for five consecutive polls (50 milliseconds), the system then determines that the die-casting cycle start signal is valid.

[0024] At this point, the system sends a trigger message to the laser barcode scanner at the entrance of the die-casting island. The scanner scans the QR code on the die-casting mold transfer tray, successfully reads it, and returns the product code of the current casting as ENG-V6-AL-98765.

[0025] The system obtains the current absolute timestamp from the underlying system clock the instant it determines that the start signal is valid. This timestamp is accurate to the millisecond level, specifically 8:15:30.125 milliseconds on June 30, 2025.

[0026] When generating the traceability index, the system executes string concatenation rules. The system extracts the fixed number DC-05 from the die-casting island, combines it with the formatted start timestamp 20250630081530125 (after removing non-numeric characters), and the last five digits of the product code 98765, ultimately generating and persistently storing the traceability index corresponding to the current cylinder block casting as DC05-20250630081530125-98765. This index will then serve as the master key for retrieving all subsequent pressure curves and temperature data for this engine block.

[0027] In this embodiment, the DBBlock data block field, Address field, and Flag field in the assembly data table are used to establish the PLC data mapping relationship corresponding to the product code. The specific implementation method is as follows: In the execution process of establishing the PLC data mapping relationship corresponding to the product code in the DBBlock data block field, Address field, and Flag bit in the assembly data table, the system establishes a precise and conflict-free data communication channel between the host computer software and the underlying industrial control hardware.

[0028] Determining the PLC data block based on the DBBlock data block field is a macro-level addressing step in data mapping. In industrial programmable logic controllers (PLCs), memory is divided into multiple independent data block areas for storing variable sets from different process stages. When establishing a communication connection, the MES-SystemServo processing module first parses the value of the DBBlock data block field in the assembly data table. This field value represents the physical base address of a specific contiguous storage area in the underlying device's memory space. The system then sends an addressing message containing this base address to the underlying device via the Industrial Ethernet protocol to lock onto the entry point of the target memory partition.

[0029] Determining the data address within a PLC data block based on the Address field is a micro-level positioning step in data mapping. After locking onto the target data block, the system further parses the Address field. This field defines the specific byte offset and bit offset of the target data item within the locked data block. The system adds the base address and the offset to calculate the absolute address of the target process parameter on the physical memory module. This two-level addressing mechanism ensures that regardless of changes in the underlying device's memory structure, as long as the mapping relationship in the assembly data table remains unchanged, the system can accurately capture the unique variable node.

[0030] Determining the read / write status of the corresponding data item based on the Flag bit enables the MES-SystemServo processing module to read or write production data according to the data mapping relationship. This is the core handshake mechanism that ensures conflict-free high-frequency communication. The Flag bit acts as the communication control word between the host computer and the underlying device. The system presets multiple levels of flag bit status codes, corresponding to specific stages such as idle, request to read, request to write, processing, and operation completed.

[0031] The Flag flag uses a status coding method to control the data interaction process. The system sets different status flags based on the data reading, data writing, and device response processes.

[0032] Specifically, when the Flag is zero, it indicates that the current communication channel is idle and new data interaction tasks are allowed; when the Flag is one, it indicates that the underlying device is updating the corresponding data area and the system temporarily suspends read operations; when the Flag is two, it indicates that the host computer has initiated a data write request and locked the current communication channel, and the underlying device is prohibited from modifying the corresponding data area; when the Flag is three, it indicates that the data write is complete and the underlying device can read the latest data content; when the Flag is four, it indicates that an abnormality has occurred in the current communication process and it is necessary to wait for the communication state to be re-established.

[0033] When executing data acquisition or control command issuance, the MES-SystemServo processing module first reads the current Flag status and only executes subsequent data access actions if the communication status meets the corresponding operation conditions. This status control method enables sequential control of data access between the host computer and the PLC, avoiding data conflicts caused by multiple data interaction tasks accessing the same data area simultaneously.

[0034] When the system needs to perform data interaction, it must verify the current status word of the Flag bit. Only when the status word indicates that the channel is idle or a specific value that allows operation will the MES-SystemServo processing module issue a read or write command along the absolute physical address determined in the aforementioned steps, and the system or underlying device will overwrite the flag bit after the command is executed.

[0035] Taking the real-time injection pressure data acquisition and process formula distribution process in the die-casting island as an example, the core control cabinet of the die-casting island adopts a Siemens series programmable logic controller. The system configures the assembly data table in the MES-SystemServo processing module. For the real-time injection pressure production data, the DBBlock data block field in the assembly data table is set to an integer value of 100, the Address field is set to a floating-point offset of 24.0, and the Flag bit is set to the communication status register M10.0.

[0036] When the system performs a read operation, it first locates the 100th data block of the controller based on the value 100 of the DBBlock data block field. Then, based on the value 24.0 of the Address field, it determines that the target data is located within the 100th data block at a starting offset of 24 bytes, and the data length occupies 4 bytes. Before initiating the read operation, the system listens to the value of the status register M10.0 over the network. The system defaults to a value of 0 for this flag indicating free memory, and a value of 1 indicating that the underlying device is refreshing data at that address.

[0037] When the system detects that the flag bit value is 0, the MES-SystemServo processing module initiates a floating-point read instruction to the 24th byte of the 100th data block of the controller according to the data mapping relationship, obtains the current injection pressure value of 85.5 MPa, and then stores the value associated with the product code.

[0038] When performing the process recipe writing operation, the standard mold temperature setting value corresponding to the current product code is sent downwards. This recipe data is set to 50 in the DBBlock data block field, 12.0 in the Address field, and the Flag bit is set to the handshake register M12.0 in the assembly data table.

[0039] The system defaults to a triple status word for the write handshake flag: 0 indicates device readiness, 2 indicates the host computer requests and locks the write operation, and 3 indicates write completion. The system first reads M12.0, confirms its value is 0, and then the MES-SystemServo processing module proactively overwrites the value of M12.0 to 2, locking the communication channel to prevent the underlying device from reading incomplete recipe data at this moment. After the channel is locked, the system continuously writes the set mold temperature value of 220 degrees Celsius starting from the 12th byte of data block 50. After the write operation is completely completed, the system overwrites the value of M12.0 to 3. Upon detecting the flag changing to 3, the underlying controller calls the latest temperature value of 220 degrees Celsius from the 12th byte of data block 50 to adjust the cooling water valve, and automatically resets M12.0 to 0 after the adjustment, awaiting the next data mapping interaction.

[0040] In this embodiment, key production data of the die-casting process are collected in real time using MODBUS-TCP. The specific implementation method is as follows: Obtaining the die-casting process status word output by the PLC and determining the current die-casting stage based on this status word is the logical prerequisite for implementing frequency conversion data acquisition. Inside the programmable logic controller (PLC) of the die-casting machine, a dedicated status register address is preset. The value in this register is not virtually generated by software, but rather an integer status word output in real time after hardware-level combinational logic operations on the signals from limit switches, displacement sensors, and pressure relays of the hydraulic system on the die-casting machine's physical body. During operation, the heterogeneous communication protocol module establishes an independent high-priority communication thread, which continuously reads the value of this status register address via the MODBUS-TCP protocol. The system internally has a status mapping table that directly maps different read integer values ​​to the corresponding physical production stages.

[0041] The first acquisition cycle is used during the mold closing and rapid injection stages, while the second acquisition cycle is used during the pressure holding, mold opening, and standby stages. The first acquisition cycle is shorter than the second acquisition cycle and is the core scheduling action for dynamically adjusting the communication load. The mold closing stage involves high-speed mold closure and high-pressure locking, while the rapid injection stage involves ultra-high-speed filling of molten metal in an extremely short time. The displacement and pressure parameters in these two stages exhibit dramatic transient nonlinear changes at the millisecond level. Therefore, when the heterogeneous communication protocol module determines that it has entered these two stages, it switches the MODBUS-TCP protocol data packet sending interval to the first acquisition cycle. The first acquisition cycle is set to an extremely small time interval threshold to ensure that the data points of the transiently changing process curves can be completely captured.

[0042] During the pressure holding phase, the hydraulic system output remains constant. During the mold opening and standby phases, the equipment is in a slow reset or stationary state, and the process parameters are basically linearly distributed or remain unchanged. When the heterogeneous communication protocol module determines that it has entered these three phases, it switches the data packet sending interval to the second acquisition cycle. The value of the second acquisition cycle is much larger than the first acquisition cycle. Internally, the heterogeneous communication protocol module uses a reloadable adaptive timer to control the initiation frequency of acquisition messages. When a status word transition triggers a die-casting stage switch, the system immediately writes a new cycle parameter into this adaptive timer. This mechanism enables the acquisition system to provide high data resolution when critical features appear, thereby proactively reducing the network request frequency during the process stabilization period, thus avoiding bandwidth congestion caused by a large number of redundant communication requests on the same Ethernet link.

[0043] Taking a die-casting island in a large die-casting workshop that produces automotive shock absorber tower parts as an example, the core controller of the die-casting island sets the address with a word offset of 4 in the DB20 data block as the output source of the die-casting process status word. The system's internal status mapping table is preset with the following rules: a value of 10 represents the standby stage, a value of 20 represents the mold closing stage, a value of 30 represents the slow injection stage, a value of 40 represents the fast injection stage, a value of 50 represents the holding pressure stage, and a value of 60 represents the mold opening stage. Simultaneously, the system presets the first acquisition cycle to 5 milliseconds and the second acquisition cycle to 100 milliseconds.

[0044] When the die-casting machine is in a no-load start state, the heterogeneous communication protocol module reads a status word value of 10 via MODBUS-TCP. The system determines it is currently in standby mode and sets the adaptive timer's timeout to 100 milliseconds. Every 100 milliseconds, the system sends a request to the controller to collect temperature and pressure data.

[0045] When the operator presses the start button, the equipment begins mold closing. The underlying displacement sensor detects the movement, and the controller updates the status word value to 20. After reading the value 20, the heterogeneous communication protocol module immediately determines that the mold closing stage has begun and instantly reloads the adaptive timer's timeout to 5 milliseconds. At this time, the system continuously and intensively collects the peak pressure data of the mold closing cylinder at a high frequency of 200 times per second.

[0046] As the process progresses, when the molten metal is injected into the mold cavity at high speed, the status word value jumps to 40. The system determines that it has entered the fast injection stage, and the timer continues to maintain the first acquisition cycle of 5 milliseconds to accurately record the pressure impact waveform at the moment the molten metal fills the mold.

[0047] After filling is complete, the cooling and solidification process begins, and the controller updates the status word to 50. The system determines that it has entered the pressure holding stage. Since the pressure holding pressure is relatively stable, the system immediately reloads the timeout of the adaptive timer back to 100 milliseconds. In the subsequent mold opening stage with a value of 60 and the standby stage with a value of 10, the system maintains a low-frequency sampling with a second acquisition cycle of 100 milliseconds until the next mold closing action triggers another change in the status word.

[0048] In this embodiment, historical data exchange is performed via OPC UA, and the specific implementation method is as follows: In the historical data exchange process via OPC UA in the heterogeneous communication protocol module, the system establishes a time-series-based local data retention and breakpoint resumption mechanism. Specifically, this step involves writing historical production data acquired via OPC UA into a local sequence cache pool according to device identifier and acquisition time. When a communication connection anomaly is detected, the incomplete historical production data is retained in the local sequence cache pool and transmitted in chronological order of acquisition time after the communication connection is restored.

[0049] Writing historical production data acquired via OPC UA into a local sequence cache pool according to device identifier and acquisition time is a physical storage prerequisite for achieving data time-series integrity. During operation, the heterogeneous communication protocol module subscribes to event and alarm information from various underlying controllers within the die-casting island via the OPC UA protocol interface. This information, triggered by specific process conditions, is collectively referred to as historical production data. The system allocates an independent, fixed-capacity, contiguous memory space, defined as the local sequence cache pool, within the local non-volatile storage medium of the edge computing gateway where the heterogeneous communication protocol module resides. This cache pool uses a circular queue data structure to manage data. When the module receives historical production data, the system extracts two core index elements from the packet header. The first element is the device identifier, representing the physical node number of the specific die-casting device that generated the data. The second element is the acquisition time, i.e., the absolute hardware timestamp accurate to milliseconds attached when the underlying hardware generated the data. The system uses the device identifier as a classification label and the acquisition time as a sorting weight to structure the historical production data and append it to the end of the local sequence cache pool.

[0050] When a communication connection anomaly is detected, incomplete historical production data is retained in a local sequence buffer and transmitted in chronological order after the communication connection is restored. This is a control action to ensure the continuity of the traceability data chain. The heterogeneous communication protocol module maintains a network status monitoring thread, which sends probe messages to the central server according to a preset heartbeat sending cycle. The system presets a heartbeat timeout threshold, which is the maximum tolerable time for the system to not receive a response message from the server. When the cumulative time the monitoring thread does not receive a response exceeds this threshold, the system determines that a communication connection anomaly has occurred. At this time, the heterogeneous communication protocol module locks the read pointer of the local sequence buffer, and newly written historical production data only accumulates within the buffer, achieving retention.

[0051] During communication link anomalies, the monitoring thread continuously sends probe messages. Once a response message is received from the central server, the system determines that the communication connection has been restored. The heterogeneous communication protocol module releases the read lock on its local sequence buffer pool. The system iterates through all historical production data with incomplete transmissions backed up in the buffer pool, extracts their collection times, and uses a quicksort algorithm to sort the backlogged data in ascending order. After sorting, the module strictly follows the chronological order of collection time, sequentially packaging and uploading the backlogged data to the central server to complete the data resuming process.

[0052] For example, in the traceability control process of an actual die-casting production line, a heterogeneous communication protocol module is deployed within an industrial edge gateway. This edge gateway has a built-in 512-megabyte solid-state storage area as a local sequence cache pool.

[0053] The core die-casting machine, Node85, within the die-casting island triggered an alarm indicating excessive mold temperature during production. The underlying hardware generated this alarm data message at 18:25:30:150 on June 30, 2025. The heterogeneous communication protocol module, after obtaining this historical production data via the OPC UA protocol, parsed out the device identifier Node85 and the acquisition time 20250630182530150. The system encapsulated this alarm content, along with the device identifier and acquisition time, into a standard structure and wrote it to the local sequence cache pool.

[0054] The system's network heartbeat sending cycle is set to 1 second, and the heartbeat timeout threshold is set to 3 seconds. At 18:25:31, the workshop core switch experienced a power outage and restart. The network status monitoring thread sent three consecutive heartbeat probe messages over the next 3 seconds, but received no acknowledgment. At 18:25:34, the system reached the heartbeat timeout threshold, determining that a communication connection anomaly had occurred.

[0055] After detecting the anomaly, the system locked the data push port to the central server. The Node85 die-casting machine generated two hydraulic system alarm data points consecutively at 18:25:35.100 and 18:25:38.200. After the module obtained these two historical production data points, it appended them to a 512-megabyte local sequence cache pool for retention. At this point, the cache pool contained a total of three untransmitted historical production data points.

[0056] At 18:25:45, the switch restarted successfully, the monitoring thread received a heartbeat response, and the system determined that the communication connection had been restored. The heterogeneous communication protocol module scanned the backlog data in the local sequence cache pool. The system compared the acquisition times of these three data entries and sorted them in ascending order according to the timestamps corresponding to 20250630182530150, 18:25:35.100, and 18:25:38.200. After the sorting was completed, the module sequentially resent these three historical production data entries to the central server.

[0057] In this embodiment, the collected production data is associated with product codes in a time sequence. The specific implementation method is as follows: In the execution flow of the die-casting island management system module, which correlates the collected production data with product codes in a time sequence, the system establishes an absolute time scale based on a unified clock source to achieve precise binding between discrete process data and specific physical castings. This step specifically includes attaching hardware timestamps to the production data, determining the data association time interval based on the start and end times of the die-casting cycle corresponding to the product code, and associating production data whose collection time falls within the data association time interval to the corresponding product code.

[0058] Adding hardware timestamps to production data is the core physical foundation for eliminating network latency and system scheduling errors. Because the die-casting island contains various heterogeneous devices and has an extremely high data acquisition frequency, software timestamps allocated by the host operating system kernel are prone to timing drift of tens of milliseconds due to CPU task queuing. To address this issue, a precise time protocol network is pre-deployed between the server housing the die-casting island management system module and the underlying programmable logic controllers and edge gateways. This protocol corrects for device clock deviations through message exchange between master and slave clocks and link latency calculations, keeping the time synchronization error of the entire workshop within the microsecond range. When the underlying hardware acquires transient pressure or temperature data and sends it through the network interface, the hardware chip of the underlying network card directly imprints a hardware timestamp based on the precise time protocol into the network packet header the instant the data frame leaves the physical port. This hardware timestamp accurately reflects the absolute physical time of the process parameters.

[0059] Determining the data association time interval based on the start and end times of the die-casting cycle corresponding to the product code is a logical step in defining the data ownership boundary. The system pre-defines specific signal mappings to identify the start and end of the die-casting cycle. When acquiring the product code of the casting at the start of the die-casting cycle, the system simultaneously records the starting hardware timestamp that triggered the code reading action as the die-casting cycle start time. As the process progresses, when the underlying equipment outputs a pre-defined cycle end signal, such as mold opening or robotic arm grasping completion, the system again captures the hardware timestamp of this signal as the die-casting cycle end time. The die-casting island management system module uses these two absolute time points as the lower and upper limits on the time axis to construct a mathematically closed interval for that specific product code, i.e., the data association time interval.

[0060] Associating production data collected within the data association time interval with the corresponding product code is the final data processing action to complete the traceability chain puzzle. The die-casting island management system module is equipped with a high-speed time-series data processing engine. When the system receives massive amounts of discrete production data with hardware timestamps, the processing engine extracts the hardware timestamp from the header of each data packet as a comparison object. The engine performs a logical AND operation to determine if the comparison object is greater than or equal to the start time of the die-casting cycle, and simultaneously less than or equal to the end time of the die-casting cycle. If the result is true, the system determines that the production data was generated within the current casting's manufacturing lifecycle and immediately adds the corresponding product code as a foreign key identifier to the record row of that production data in the database. If the result is false, the data belongs to the previous or next die-casting cycle, and the system refuses to bind it to the current product code.

[0061] For example, taking a die-casting production process as an example, all node devices in the system network have enabled the precise time protocol, and the master-slave clock deviation is stable within 1 microsecond.

[0062] We are currently entering the battery casing production cycle numbered 56789. The bottom mold closing sensor was triggered at 10:15:20:100 milliseconds on June 30, 2025. The die-casting island management system module received the cycle start signal and bound the read product code 56789 with the start time of 10:15:20:100 milliseconds on June 30, 2025.

[0063] After the melting, injection, and cooling processes, the bottom ejector pin retraction sensor was triggered at 10:16:50.400 milliseconds on June 30, 2025. The system used this as the cycle end signal and recorded the die-casting cycle end time. Therefore, the data association time interval constructed by the system for product code 56789 is from 10:15:20.10 milliseconds on June 30, 2025 to 10:16:50.400 milliseconds on June 30, 2025.

[0064] During production, the underlying controller continuously reports pressure data from the injection cylinders. One data packet with a pressure value of 120.5 MPa has a hardware timestamp tagged by the network card at 10:15:45.200 milliseconds on June 30, 2025. The system's timing processing engine extracts this timestamp and finds that it falls within the data association time interval defined above. Subsequently, the system associates the 120.5 MPa pressure record with product code 56789 in the database.

[0065] Subsequently, the system received another hydraulic pump cooling status data, with a hardware timestamp of 10:16:51:000 milliseconds on June 30, 2025. The system comparison revealed that this timestamp was greater than the currently set die-casting cycle end time, and determined that it was outside the range. Therefore, this cooling status data was not associated with product code 56789.

[0066] In this embodiment, historical process data is backtracked when a fault alarm is triggered during quality control. The specific implementation method is as follows: In the execution process of the die-casting island management system module, when a fault alarm is triggered by quality control, the system backtracks historical process data. It abandons the traditional fixed-duration waveform recording method and establishes a dynamic sliding window traceability engine based on physical parameter attributes. This step specifically includes obtaining the trigger time of the fault alarm and determining the historical data backtracking time window based on the corresponding process parameters. When the mold temperature corresponding to the fault alarm is abnormal, the time window is extended after the fault alarm trigger time; when the injection pressure corresponding to the fault alarm is abnormal, the time window is shortened.

[0067] To ensure different data backtracking ranges correspond to different types of process anomalies, the die-casting island management system module pre-establishes a mapping relationship between process parameter types and time window adjustment parameters. The system calls the corresponding time window adjustment parameters based on the process parameter category corresponding to the fault alarm, adjusting the historical data backtracking range accordingly.

[0068] Specifically, the system configures corresponding data backtracking strategies for different types of faults, such as abnormal mold temperature, abnormal injection pressure, abnormal hydraulic system, and abnormal equipment operation. For abnormal mold temperature, which is characterized by heat conduction hysteresis, the system increases the data retention ratio after the fault trigger time to fully capture the response process following the temperature change. For abnormal injection pressure, which is a transient impact type of fault, the system reduces the data retention range before and after the incident, concentrating the backtracking data on the pressure change phase. For abnormal hydraulic system, the backtracking range is adjusted based on the duration of the pressure change. For abnormal equipment operation, the corresponding data extraction range is determined based on the length of the operation cycle.

[0069] The system maintains a mapping table between anomaly types and time window adjustment parameters. Upon detecting a fault alarm, it parses the corresponding process parameter type based on the alarm code and calls the corresponding time window parameters to filter historical process data. The time window adjustment parameters are determined by the die-casting equipment's operating cycle, the rate of change of process parameters, and historical fault sample data, and are updated during production based on changes in equipment status.

[0070] Obtaining the trigger time of a fault alarm and determining the historical data backtracking time window based on the corresponding process parameters is the fundamental scheduling step for achieving dynamic and accurate traceability. The die-casting island management system module maintains a process parameter time window mapping table. For various fault alarms reported by the equipment, the system first captures the absolute hardware timestamp of the alarm message's generation and defines it as the fault alarm's trigger time. The system then looks up the table to obtain the specific process parameter category corresponding to the alarm. The system sets a standard data backtracking time window for regular process parameters. This standard time window starts from the trigger time minus a standard pre-offset and ends with the trigger time plus a standard post-offset.

[0071] When a fault alarm corresponds to an abnormal mold temperature, extending the time window after the alarm trigger time is a targeted control action based on the thermodynamic inertial physical law. Changes in mold temperature depend on the heat conduction of the mold steel and the heat exchange of the cooling water network; this physical process has a strong hysteresis effect. When the system detects an excessive mold temperature triggering an alarm, the abnormally accumulated heat often hasn't been fully conducted to the sensor surface, and the intervention actions of the underlying control system, such as activating the backup cooling water circuit or adjusting the flow valve, require a considerable amount of time for the temperature curve to converge again. If a standard time window is used, the captured data curve is easily truncated midway when the temperature is still rising or hasn't subsided. Therefore, when the system determines the process parameter to be mold temperature, it calls the time window extension algorithm, significantly increasing the value of the post-time offset. The system extends the end time of the backtracking to after the alarm trigger time to ensure that the entire thermodynamic evolution curve from the abnormal temperature development and peak to the effective cooling intervention is captured.

[0072] When a fault alarm corresponds to abnormal injection pressure, shortening the time window is a precise data filtering action based on the transient characteristics of hydraulic fluid dynamics. Pressure anomalies in the injection cylinder typically manifest as excessively high impact peaks within a very short time or a sudden pressure collapse during the holding period. Hydraulic waves propagate at the speed of sound in the pipeline, making pressure anomalies high-frequency transient events on the millisecond scale. Using excessively long time windows for data extraction would not only introduce a large amount of irrelevant steady-state pressure data during the stable period, interfering with root cause analysis, but also result in an excessively large trace data package size. Therefore, when the system determines the process parameter to be injection pressure, it invokes a time window contraction algorithm to simultaneously reduce the values ​​of the preceding and following time offsets. The system narrows the time window to its limit, stripping away redundant steady-state data and directly focusing on the ultra-high-frequency waveform cross-section within tens of milliseconds of pressure surge or drop.

[0073] For example, the standard forward offset in the die-casting island management system module is set to 5 seconds, and the standard backward offset is set to 5 seconds.

[0074] During a specific die-casting cycle, the underlying equipment reported an alarm indicating that the temperature in the moving mold area exceeded the limit at 14:15:20:000 on June 30, 2025. After obtaining the trigger time, the system identified the corresponding mold temperature parameters. The system then invoked a time window extension algorithm, doubling the subsequent time offset from the standard value of 5 seconds to 30 seconds, while keeping the preceding offset unchanged at 5 seconds. The system calculated the historical data backtracking time window for this temperature anomaly to start at 14:15:15:000 and extend to 14:15:50:000. The system ultimately extracted this 35-second period of historical process data from the temperature sensor and cooling water flow valves. This not only captured the heat accumulation process before the anomaly occurred but also completely recorded the temperature drop response curve within 30 seconds after the alarm was triggered, with the cooling water valves fully open, providing complete closed-loop evidence for analyzing whether the mold water circuit was clogged or scaled.

[0075] In subsequent production, the underlying equipment reported an abnormal peak pressure at the end of the rapid injection at 15:20:10:500 on June 30, 2025. The system obtained the trigger time and identified the parameter as injection pressure. The system immediately invoked the time window contraction algorithm, shortening the preceding time offset from the standard value of 5 seconds to 0.5 seconds, and the following time offset from the standard value of 5 seconds to 0.5 seconds. The system calculated the start of the time window to be 15:20:10:000 milliseconds and the end point to be 15:20:11:000 milliseconds. The system extracted only the high-frequency pressure time sequence within this 1-second window from the database, accurately capturing the core interval representing the transient waveform of hydraulic impact, thus reducing the computational load on the system when accessing historical pressure data.

[0076] In this embodiment, the traceable transient process data is bound to the abnormal product code to generate a traceability record. The specific implementation method is as follows: In the execution process of binding the traceable transient process data with the abnormal product code to generate traceability records in the system's main control screen module, the system establishes a tamper-proof data solidification mechanism based on digital fingerprint technology to meet the compliance audit requirements of high-standard industrial manufacturing. This step specifically includes acquiring transient process data, abnormal product codes, and the corresponding time of occurrence of the anomaly; calculating a digest of the transient process data, abnormal product codes, and anomaly occurrence time based on a hash digest algorithm to generate a corresponding traceability digest value; and then associating and storing the traceability digest value with the traceability record.

[0077] Acquiring transient process data, abnormal product codes, and corresponding anomaly occurrence times is a preprocessing step in preparing the encrypted information source. When the quality control module triggers a fault alarm and completes time window backtracking, the background service program of the system's main control screen module retrieves the raw payloads of these three core data dimensions from memory. To ensure absolute consistency and reproducibility of the data used in the summary calculation, the system first performs a standardized format conversion on the transient process data array obtained from the backtracking. The system serializes the two-dimensional process data array, containing multiple sets of timestamps and floating-point values, into a standard data exchange text format according to fixed key-value pair rules, while simultaneously unifying the text's character encoding standard.

[0078] The core cryptographic action for generating a unique digital fingerprint of data is to calculate the digest of transient process data, abnormal product codes, and the time of anomaly occurrence using a hash digest algorithm. The system backend is equipped with a dedicated hash function calculation engine. The engine first concatenates the string of the abnormal product code, the absolute timestamp string of the anomaly occurrence time, and the text string of the transient process data generated in the previous serialization step into a very long raw plaintext string, according to preset concatenation rules. Then, the system calls a standard hash digest algorithm such as SHA-256 to perform a hash operation on this very long raw plaintext string. This algorithm maps input plaintext of arbitrary length to a fixed-length output hash value. Regardless of whether the input process data curve contains hundreds or thousands of sampling points, or only a few, the system will ultimately generate a unique, collision-resistant, and irreversible fixed-length hexadecimal string, which is the generated traceability digest value.

[0079] Associating traceability summary values ​​with traceability records is a crucial step in achieving deep binding and physical solidification of multi-dimensional data. The system's main control panel creates a dedicated underlying data table for traceability records within a relational database. This table includes not only product code fields, failure time fields, and binary large object fields storing historical process parameters, but also a dedicated independent summary authentication field for storing digital fingerprints. The system directly writes the fixed-length hexadecimal traceability summary value generated in the previous step into the summary authentication field of this record row. At this point, transient process parameters and specific abnormal castings are tightly bound together by encryption algorithms at both the logical and data storage levels. Any attempt to modify temperature or pressure values ​​in the database will result in a mismatch between the recalculated summary value and the original summary value stored in that field, thus rendering any tampering impossible.

[0080] For example, in the background server of the system's main control screen module, the SHA-256 hash digest algorithm is used by default, and the data splicing order is specified as abnormal product code, abnormal occurrence time and transient process data, and all text is encoded in UTF-8.

[0081] A die-casting machine triggered a rapid injection anomaly during production. The system's main control screen module obtained the current anomaly product code as AL-998877, with an absolute timestamp corresponding to the anomaly occurrence time of 20250630164520500. Simultaneously, the system retrieved the transient injection pressure sequence before and after this time point through time window backtracking. The system converted this pressure sequence into a standard JSON format text stream, the text content being a serialized string containing the sampling time point and three consecutive MPa pressure values ​​of 150.5, 120.2, and 80.5 MPa.

[0082] The system data conversion engine seamlessly combines the product code AL-998877, the timestamp 20250630164520500, and the aforementioned JSON-formatted pressure string into a single long plaintext string.

[0083] Subsequently, the system invokes the SHA-256 algorithm engine to perform one-way hashing on the long plaintext string. After multiple iterations of bitwise operations and nonlinear functions, the engine outputs a 64-bit hexadecimal traceback digest value. The specific content of this digest value is e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855.

[0084] Upon receiving the traceability summary value, the system immediately performs a structured query language write operation through the database write middleware. The system inserts a new record into the central traceability database table, writing the product code AL-998877 into the primary key field, the abnormal time and stress data in JSON format into the business field, and simultaneously writing the 64-bit hexadecimal traceability summary value directly into the authentication-specific field named HashAuth. Through this field's storage on disk, the system establishes a permanent and unforgeable digital file for this defective aluminum alloy die-cast part.

[0085] In this embodiment, the status of the LED signboard used for scanning is updated based on the alarm signal linkage. The specific implementation method is as follows: In the execution process of updating the LED signboard status for scanning based on alarm signals in the system's main control screen module, the system establishes a hardware-software decoupled display control mechanism based on a publish-subscribe model to achieve physical workshop visualization of abnormal states. This step specifically includes converting fault alarms into corresponding signboard status information and sending this information to the LED signboard interface via an asynchronous message event bus, enabling the LED signboard interface to update its display content based on the received status information.

[0086] Converting fault alarms into corresponding Kanban status information is a fundamental preprocessing step for achieving cross-device data semantic unification. Fault alarms reported by the die-casting island's underlying equipment are typically structured messages containing fault codes, timestamps, and original parameter values. LED Kanban devices cannot directly parse such industrial control messages. Therefore, the system's main control screen module has a pre-built status mapping data table. When the module receives a fault alarm, it extracts the fault code from the message as the primary key and performs a matching query in the status mapping data table. If the query matches, the system extracts three preset display control parameters: the text string used for presentation, the values ​​of the red, green, and blue primary color channels that determine the display color, and the time interval parameter controlling the text blinking frequency. The system encapsulates these three display control parameters into a standard data payload, which is the Kanban status information.

[0087] Sending Kanban status information to the LED Kanban interface via an asynchronous message event bus is a crucial communication architecture action to ensure that the core control logic is not blocked by network latency from external devices. In the background runtime environment of the system's main control screen module, the main business logic thread and the Kanban communication thread are completely physically isolated. A memory-based data buffer channel, namely the asynchronous message event bus, is established between them. The main business logic thread of the system's main control screen module, acting as the message publisher, only needs to push the data payload into a pre-defined queue of specific publishing topics within the event bus after generating Kanban status information, and then immediately returns to continue processing other high-frequency production data from the die-casting machine. The main business logic thread does not wait for any reception confirmation from the LED Kanban interface, thus achieving complete time decoupling between the system's core acquisition process and the external display hardware.

[0088] The LED signboard interface updates its display content based on received signboard status information, which is the final hardware rendering action for visualizing abnormal states. As a message subscriber to the event bus, the LED signboard interface is configured with an independent background polling thread. This thread continuously listens to the specific published topic queue at a fixed clock cycle. When new signboard status information is detected in the queue, the LED signboard interface actively pulls it into local memory and performs deserialization and parsing. The interface program generates dot matrix font data based on the parsed text string, sets the duty cycle of each pixel's LED based on the red, green, and blue primary color channel values, and generates underlying refresh pulse signals according to time interval parameters. Finally, the interface program sends these underlying physical control commands to the LED signboard's hardware driver control board via a serial communication port or Ethernet port, driving the LED matrix to change its state.

[0089] For example, the status mapping data table configured in the system's main control screen module defines the following conversion rules. For the fault alarm with code E404, the corresponding text string is "Mold water temperature exceeds limit," and the values ​​for the red, green, and blue primary color channels are set to red channel 255, green channel 0, and blue channel 0. The time interval parameter for the text flashing frequency is set to 500 milliseconds. Here, red channel 255 represents the red LED reaching maximum brightness, and the time interval of 500 milliseconds means the text switches between bright and dark every 0.5 seconds to create a warning flashing effect.

[0090] The system's asynchronous message event bus allocates a circular queue in memory with a capacity of 1024 messages, and defines the publication topic name as BoardUpdate. The polling thread cycle for the LED signboard interface is set to 50 milliseconds. The physical LED signboards deployed on the die-casting island have a resolution of 192 by 64 pixels.

[0091] At 17:20:15 on June 30, 2025 (absolute time), the quality control system detected that the mold temperature reached 235 degrees Celsius and sent a fault alarm with fault code E404 to the main control screen module of the system.

[0092] The system's main control screen module extracts fault code E404, retrieves the status mapping data table, and generates a Kanban status information data payload containing text indicating that the mold water temperature exceeds the limit, a red channel color value of 255, and a 500-millisecond flashing interval. The main business logic thread pushes this data payload into the BoardUpdate release topic queue of the event bus in 2 milliseconds, and then immediately ends the current processing cycle to continue listening for data from the next production cycle of the die-casting machine.

[0093] Thirty milliseconds later, the polling thread of the LED display interface detected a new payload in the BoardUpdate queue. The interface program retrieved and parsed it, extracting the text "Mold water temperature exceeds limit" and the red flashing instruction. The interface program called the system's internal font library, converted the text into a 192x64 resolution dot matrix two-dimensional array, and sent the dot matrix array and color pulse signal to the LED hardware driver board on-site through port 502 of the transmission control protocol. The LED display on-site immediately displayed the words "Mold water temperature exceeds limit" in red across the entire screen, and flashed at a high frequency once every 500 milliseconds, prompting on-site inspection personnel to immediately go to the die-casting island for physical intervention.

[0094] The above description is merely a preferred embodiment of the present invention, and the scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for users of ordinary technical skills, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A traceability control system for die-casting islands in high-pressure die casting, characterized in that, include: The MES-SystemServo processing module is configured to acquire the product code of the casting at the beginning of the die casting cycle and create an assembly data table for data acquisition and processing. The assembly data table includes a DBBlock data block field for underlying mapping, an Address field, and a Flag bit that controls the read / write status. The heterogeneous communication protocol module integrates OPC UA and MODBUS-TCP protocols; based on the Flag bit, it uses the DBBlock and Address fields to address the PLC, collects key production data of the die-casting process in real time through MODBUS-TCP, and exchanges historical data through OPC UA. The die-casting island management system module includes multiple integrated subsystem interfaces, configured to sequentially associate the collected production data with the product code; When a fault alarm is triggered during quality control, the abnormal time point is extracted and the historical process data before and after that time point is traced back. The system's main control screen module, serving as a comprehensive entry point for communication and control, is configured to bind retrospective transient process data with abnormal product codes to generate traceability records, and to update the status of the LED dashboard used for scanning based on alarm signals.

2. The high-pressure die-casting island casting traceability control system according to claim 1, characterized in that, The MES-SystemServo processing module obtains the product code of the casting at the beginning of the die-casting cycle, specifically including: After detecting the die casting cycle start signal output by the die casting equipment, the product code of the current casting is read, and a correspondence is established between the product code and the start time of the current die casting cycle to generate a traceability index corresponding to the current casting.

3. The high-pressure die-casting island casting traceability control system according to claim 1, characterized in that, The DBBlock data block field, Address field, and Flag field in the assembly data table are used to establish the PLC data mapping relationship corresponding to the product code, specifically including: The PLC data block is determined based on the DBBlock data block field, the data address in the PLC data block is determined based on the Address field, and the read / write status of the corresponding data item is determined based on the Flag bit, so that the MES-SystemServo processing module can read or write production data according to the data mapping relationship.

4. The high-pressure die-casting island casting traceability control system according to claim 1, characterized in that, The heterogeneous communication protocol module collects key production data of the die-casting process in real time via MODBUS-TCP, specifically including: Obtain the die-casting process status word output by the PLC, and determine the current die-casting stage based on the die-casting process status word; The first acquisition cycle is used during the mold closing and fast injection stages, and the second acquisition cycle is used during the pressure holding, mold opening and standby stages. The first acquisition cycle is shorter than the second acquisition cycle.

5. The high-pressure die-casting island casting traceability control system according to claim 1, characterized in that, The heterogeneous communication protocol module exchanges historical data via OPC UA, specifically including: Historical production data obtained through OPC UA is written to the local sequence cache pool according to the device identifier and collection time; When a communication connection anomaly is detected, the historical production data that has not been transmitted is retained in the local sequence cache pool and transmitted in the order of the collection time after the communication connection is restored.

6. The high-pressure die-casting island casting traceability control system according to claim 1, characterized in that, The die-casting island management system module performs a time-series association between the collected production data and the product code, specifically including: A hardware timestamp is added to the production data, and the data association time interval is determined by the start time and end time of the die-casting cycle corresponding to the product code. Production data whose collection time is within the data association time interval is associated with the corresponding product code.

7. The high-pressure die-casting island casting traceability control system according to claim 1, characterized in that, When the die-casting island management system module triggers a fault alarm during quality control, it backs up the historical process data, specifically including: Obtain the trigger time of the fault alarm and determine the historical data backtracking time window based on the process parameters corresponding to the fault alarm; When the fault alarm corresponds to an abnormal mold temperature, the time window is extended after the fault alarm trigger time; When the fault alarm corresponds to an abnormal injection pressure, the time window is shortened.

8. The high-pressure die-casting island casting traceability control system according to claim 1, characterized in that, The system's main control screen module binds the retrospective transient process data with the abnormal product codes to generate traceability records, specifically including: The transient process data, abnormal product code, and corresponding abnormal occurrence time are obtained. A hash digest algorithm is used to perform a digest calculation on the transient process data, abnormal product code, and abnormal occurrence time to generate a corresponding traceability digest value. The traceability digest value is then associated with and stored with the traceability record.

9. The high-pressure die-casting island casting traceability control system according to claim 1, characterized in that, The system's main control screen module updates the status of the LED signboard used for scanning based on alarm signals, specifically including: The fault alarm is converted into corresponding dashboard status information, and the dashboard status information is sent to the LED dashboard interface via an asynchronous message event bus, so that the LED dashboard interface updates the displayed content according to the received dashboard status information.

10. A method for traceability control of die-casting island castings in high-pressure casting, implemented by a traceability control system for die-casting island castings according to any one of claims 1 to 9, characterized in that... include: S1: At the start of the die casting cycle, obtain the product code of the casting and create an assembly data table for data acquisition and processing; The assembly data table includes a DBBlock data block field for underlying mapping, an Address field, and a Flag bit that controls the read / write status. S2: Based on the Flag bit, the PLC is addressed using the DBBlock data block field and the Address field. Key production data of the die-casting process are collected in real time through the integrated MODBUS-TCP protocol, and historical data is exchanged through the integrated OPC UA protocol. S3: Perform a time-series correlation between the collected production data and the product code; When a fault alarm is triggered during quality control, the abnormal time point is extracted and the historical process data before and after that time point is traced back. S4: Bind the traced transient process data with the abnormal product code to generate a traceability record, and update the status of the LED Kanban used for scanning based on the alarm signal.

Images