Injection molding raw material feeding mistake-proofing checking system for safety system air bag

Abstract

The invention relates to the technical field of injection molding production and manufacturing, and discloses an injection molding raw material feeding mistake-proofing verification system for a safety system airbag, which comprises a central verification server, an edge computing gateway, an intelligent hopper unit provided with an electromagnetic locking mechanism and a handheld intelligent terminal. The handheld terminal is used for collecting information, and the server verifies order matching and generates an unlocking instruction when the physical position identifiers are consistent and are in a preset time window. The system establishes a virtual session to lock the injection molding machine in a single-batch mode, and the product taking-out manipulator and the label printer are linked in real time through the interlocking control module according to the session state. The algorithm engine module calculates single-mode consumption based on screw metering parameters and melt PVT characteristics so as to invert real-time inventory, and pressure and speed characteristics in the glue injection process are extracted in combination with the rheological analysis module to generate rheological fingerprints for attribute verification. According to the invention, the space-time physical interlocking of the feeding authority, the whole-process batch anchoring and the online closed-loop monitoring of the physical attributes of the raw materials are realized.

Description

Technical Field

[0001] This invention relates to the field of injection molding manufacturing technology, specifically to an injection molding raw material feeding error prevention and verification system for airbags in safety systems. Background Technology

[0002] Currently, raw material management in injection molding workshops largely relies on manual verification of paper documents and barcode scanning equipment. This method focuses on logical comparison at the information level. After operators scan the raw material barcode using handheld terminals, the system determines that the information matches and assumes the feeding action is compliant, lacking physical constraints on the actual feeding behavior. Under the current production model, the barcode scanning operation and the actual feeding action are not strictly linked in time and space. There is a gap between the moment the scan verification is successful and the moment the hopper lid is opened. The lack of physical locking and temporal consistency verification mechanisms leads to the risk that operators may mistakenly feed raw materials into the wrong machine's hopper even after scanning the correct label.

[0003] In continuous production scenarios involving multiple batches of raw materials, existing control measures lack strict anchoring for raw material batches within a single production order. When the batch numbers of two consecutive packages of raw materials are inconsistent, the system often lacks an effective isolation strategy, easily leading to mixing of different batches of raw materials in the barrel. Furthermore, injection molding machine operation and back-end auxiliary equipment are usually controlled independently. When an error occurs during the raw material feeding stage, the system cannot coordinate to stop the robotic arm from picking up parts and the label printing equipment, resulting in injection molded parts produced using incorrect raw materials still being retrieved, labeled, and flowing into subsequent processes, causing the product quality traceability chain to fail.

[0004] Existing inventory management methods are mostly based on standard quota deductions, which cannot reflect the actual consumption fluctuations during the injection molding machine's production process. There is a lack of calculation models that invert actual consumption based on injection molding process parameters, making it difficult for the system to detect unauthorized material feeding and raw material loss without scanning records. Current error prevention mechanisms only compare information on outer packaging labels, failing to monitor the physical properties of raw materials within the injection molding machine. When the outer packaging label is correct but the actual material inside has degraded or is of mixed types, there is a lack of online identification methods based on melt rheological characteristics, making it impossible to promptly intercept substandard raw materials from entering the molding stage. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an injection molding raw material feeding error prevention and verification system for airbags in safety systems. This system solves the problems of existing injection molding production processes relying on manual verification of information, lacking mandatory interlocking in terms of physical space and time, lacking full-process batch anchoring and back-end equipment linkage control for multi-batch production, and difficulty in inverting actual consumption based on process data and monitoring the consistency of raw material physical properties in real time.

[0006] To achieve the above objectives, the present invention provides the following technical solution: The first aspect of the present invention provides an injection molding raw material feeding error prevention and verification system for airbags in safety systems. This system includes a central verification server, an edge computing gateway, an injection molding machine unit, an intelligent hopper unit, and a handheld intelligent terminal. The central verification server, as the core data processing center, stores production orders, bills of materials, and rheological fingerprint models, and is configured to execute verification logic and generate control instructions. The edge computing gateway is physically connected to the injection molding machine unit controller and the intelligent hopper unit, configured to collect injection molding machine operating process parameters and drive the action execution components of the intelligent hopper unit. The intelligent hopper unit is equipped with an electromagnetic locking mechanism, an open / closed status detection sensor, and a near-field communication reader / writer to construct a physical-level feeding permission control checkpoint. The handheld intelligent terminal is configured to scan and identify equipment identifiers and raw material information, and serves as a human-machine interaction medium.

[0007] In the aforementioned system, this invention employs a material feeding permission control mechanism based on dual spatiotemporal constraints. A handheld smart terminal collects the device identifier and raw material batch barcode, and a central verification server performs a logical comparison based on the production order and bill of materials. After the logical comparison passes, the system does not directly grant feeding permission but instead requires the handheld smart terminal to further read the physical location identifier of the smart hopper unit. The central verification server compares the location identifier with the target device code, generating an unlocking authorization token with a time window restriction only when they match. The edge computing gateway receives this token and drives the electromagnetic locking mechanism to perform the unlocking action within the limited time window. If the open / closed state detection sensor does not detect an open signal within the time window, the edge computing gateway forcibly resets the locked state. This mechanism, through the dual binding of physical location and valid time, prevents operators from performing feeding actions on the wrong machine.

[0008] This invention further proposes a batch anchoring and consistency control method based on virtual sessions. When the session management module of the central verification server detects a first batch feeding request, it instantiates a virtual session object containing a session sequence number and an anchored batch number, and locks the injection molding machine unit into a single batch production mode. Subsequent batch feeding requests no longer perform basic bill of materials matching, but instead perform strict anchored batch comparison, allowing only raw materials with completely identical anchored batch numbers to be fed. Furthermore, the system establishes a connection with the downstream execution unit through an interlock control module. When the virtual session is active and normal, the system sends enable signals to the product removal robot and the finished product label printer; when the session ends or an anomaly occurs, the system cuts off the enable signals, physically blocking the flow of injection molded parts to the qualified product area and prohibiting the printing of traceability labels, thereby achieving full-process batch alignment from raw material input to finished product output.

[0009] To address the challenge of monitoring raw material loss or illegal mixing after feeding, this invention employs a real-time inventory hedging and anomaly detection mechanism based on a mechanistic model. The algorithm engine module acquires real-time data on the injection molding machine screw's metering start position, injection termination position, melt temperature, and back pressure via an edge computing gateway. Based on the screw's geometric parameters and the raw material melt's PVT (pressure-specific volume-temperature) characteristics, the system performs an integral calculation on the actual raw material consumption mass for each injection cycle. The system maintains a virtual inventory register, which is updated in real-time based on the feeding quantity and single-mold consumption. The system logically compares the calculated theoretical remaining inventory with the physical material shortage signal from the intelligent hopper unit: if the physical sensor indicates a material shortage but the theoretical inventory still has a surplus, it is determined to be a material loss anomaly; if the theoretical inventory is depleted but the physical sensor shows there is still material, it is determined to be a non-recorded feeding anomaly.

[0010] To address the risk of variations or incorrect material selection due to variations in the physical properties of raw materials, this invention also provides a closed-loop method for secondary verification of material properties based on rheological fingerprints. This method considers the physical retention volume from the hopper to the injection molding machine nozzle. The rheological analysis module calculates the hysteresis period required for newly added raw materials to reach the nozzle based on the remaining inventory and the consumption per mold cycle, and sets a silent counter to skip the mixing transition zone. Within the subsequent effective sampling window, the system frequently acquires real-time pressure and screw speed curves during the injection molding process. Based on the effective length of the screw metering section and the shear thinning calibration coefficient, the system performs numerical calculations on the sampled data and extracts the apparent viscosity index, which characterizes the flow resistance properties of the raw material, as the rheological fingerprint.

[0011] The system compares the measured rheological fingerprint with a pre-stored standard fingerprint template and calculates the relative deviation rate. If the relative deviation rate exceeds the preset quality control threshold, the system determines that the physical properties of the raw material are abnormal, immediately triggers an alarm, and sends a periodic prohibition command to the injection molding machine controller, forcing the equipment to stop.

[0012] Furthermore, this invention also includes a transaction cancellation processing logic based on physical action backtracking. When a cancellation request for material feeding is received, the system retrieves the opening and closing status detection sensor logs within the corresponding time window. If the logs show that the hopper cover has always remained closed, the system performs inventory rollback at the logic level; if the logs show that the hopper cover has been opened, the system determines that the raw material has been physically transferred, refuses direct cancellation, and forces the injection molding machine into a cleaning lock mode until the mass of the discharged melt reaches the preset minimum emptying and cleaning volume.

[0013] In summary, this invention achieves error prevention, anti-mixing, and closed-loop quality control in the raw material feeding process of injection molding by constructing a multi-dimensional protection system that includes physical interlocking, virtual session anchoring, process data inversion inventory, and rheological characteristic fingerprint verification.

[0014] This invention provides a material feeding error prevention and verification system for injection molding raw materials used in safety system airbags. It has the following beneficial effects: 1. This invention, through the collaboration of a central verification server and an edge computing gateway, utilizes an electromagnetic locking mechanism and a location identification verification mechanism to set an effective time window after generating a feeding instruction. The hopper is only unlocked when the physical location matches and the time is met. This method establishes a mandatory association between feeding permissions and physical space and time dimensions, eliminating the risk of operators accidentally feeding materials after scanning in a different location, and ensuring the uniqueness and accuracy of the physical path of raw material input.

[0015] 2. This invention establishes batch anchoring logic based on virtual sessions to lock the injection molding machine unit into a single raw material batch mode. The interlock control module then links the downstream product removal robot and label printing equipment in real time. When an abnormality is detected in the end of the feeding session, this method automatically cuts off the enable signal of the downstream execution unit, physically blocks the flow of injection molded parts to the qualified product area, and prohibits label printing. This achieves full-process batch alignment from raw material input to finished product output, preventing product quality traceability from being broken due to the mixing of raw materials from different batches.

[0016] 3. This invention constructs a single-mold consumption calculation model based on the injection molding machine screw action parameters and melt pressure, specific volume, and temperature characteristics through an algorithm engine module. It also extracts pressure and velocity characteristics during the injection process using a rheological analysis module to generate a rheological fingerprint. This method enables real-time offsetting and comparison between virtual inventory and physical sensor status, as well as online secondary verification of the physical flow properties of raw materials. Based on the verification of the raw material outer packaging barcode, it can further identify risks of material loss, unrecorded material input, and variations in the physical properties of raw materials, forming a quality closed loop with dual verification in the digital and physical worlds. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the hardware architecture of the injection molding raw material feeding error prevention and verification system for airbags in a safety system according to the present invention. Figure 2 This is a flowchart of the task binding and information flow verification process of the present invention.

[0018] Among them, 100 is the central verification server; 102 is the communication interface module; 110 is the order management module; 115 is the location verification module; 120 is the session management module; 130 is the interlock control module; 140 is the algorithm engine module; 150 is the transaction rollback module; 160 is the rheological analysis module; 170 is the data persistence module; 200 is the edge computing gateway; 210 is the protocol conversion module; 220 is the data cleaning module; 230 is the uplink communication module; and 300 is the injection molding machine. Units; 310, Controller; 400, Intelligent Hopper Unit; 401, Hopper Body; 402, Hopper Cover; 410, Electromagnetic Locking Mechanism; 420, Opening / Closing Status Detection Sensor; 430, Near-Field Communication Reader / Writer; 440, Local Signal Junction Box; 500, Handheld Intelligent Terminal; 510, Scanning Verification Module; 520, Data Interface Module; 600, Post-Processing Execution Unit; 610, Product Removal Robot; 620, Conveyor Belt Control Box; 630, Finished Product Label Printer. Detailed Implementation

[0019] The technical solutions in 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.

[0020] See attached document Figure 1 The present invention provides an injection molding material feeding error prevention and verification system for airbags in safety systems, comprising: a central verification server 100, an edge computing gateway 200, an injection molding machine unit 300, an intelligent hopper unit 400, and a handheld intelligent terminal 500.

[0021] The central verification server 100 is configured as the core data processing center of the system. The central verification server 100 stores a production order database, a bill of materials database, and a rheological fingerprint model library. The central verification server 100 establishes a bidirectional data communication connection with the edge computing gateway 200 via an industrial Ethernet network. The central verification server 100 also establishes a wireless data communication connection with the handheld smart terminal 500 via a wireless local area network.

[0022] The edge computing gateway 200, acting as a data acquisition and preprocessing node, is physically connected to the controller of the injection molding machine unit 300. The edge computing gateway 200 is equipped with a fieldbus communication interface. Through this interface, the edge computing gateway 200 reads data from the internal registers of the injection molding machine unit 300's controller in real time. This internal register data includes screw position data, screw torque data, injection pressure data, screw speed data, and the current mold cycle's operating status data.

[0023] Injection molding machine unit 300 is a physical device for performing plastic molding processes. Injection molding machine unit 300 includes a plasticizing component, an injection component, and a mold closing component. The plasticizing component is used to heat and melt solid granular raw materials into a fluid state. The injection component is used to inject the molten raw materials into the mold cavity.

[0024] The intelligent hopper unit 400 is installed at the feed inlet of the injection molding machine unit 300. The intelligent hopper unit 400 physically includes an electromagnetic locking mechanism 410, an open / closed status detection sensor 420, and a near-field communication reader / writer 430. The electromagnetic locking mechanism 410 is mechanically connected to the hopper cover of the intelligent hopper unit 400. The electromagnetic locking mechanism 410 is configured to perform locking and unlocking actions in response to control signals sent by the edge computing gateway 200.

[0025] An open / closed state detection sensor 420 is mounted on the edge of the hopper lid of the intelligent hopper unit 400. The sensor 420 is configured to detect the physical open and closed states of the hopper lid in real time and transmit the detected state signals to the edge computing gateway 200. A near-field communication reader 430 is mounted on the outer wall of the intelligent hopper unit 400 at an easily accessible location. The reader 430 is configured to transmit radio frequency signals and read the response signals from the handheld smart terminal 500 to confirm physical proximity events.

[0026] The handheld smart terminal 500 is configured as a mobile human-machine interface device. It integrates a barcode scanning module, a near-field communication module, and a display interaction module. The barcode scanning module is configured to optically recognize the two-dimensional barcode information on the raw material packaging bag. The near-field communication module is configured to establish a physical location binding relationship with the near-field communication reader 430 on the intelligent hopper unit 400 through close-range handshake communication. The display interaction module is configured to display operation guidance information, alarm blocking information, and a work confirmation interface.

[0027] See attached document Figure 1 The hardware structure of the intelligent hopper unit 400 further includes: hopper body 401, hopper cover 402, electromagnetic locking mechanism 410, open / closed state detection sensor 420, near-field communication reader / writer 430, and local signal junction box 440.

[0028] The hopper body 401 is constructed as a hollow container with an upper opening and a lower outlet. The lower outlet of the hopper body 401 is physically connected via a flange to the feed throat of the plasticizing component of the injection molding machine unit 300. The hopper body 401 is used to temporarily store injection molding granules to be processed. The hopper cover 402 is hinged to the edge of the upper opening of the hopper body 401. The hopper cover 402 is configured to rotate about the hinge axis, thereby switching between a closed position that closes the upper opening of the hopper body 401 and an open position that exposes the upper opening of the hopper body 401.

[0029] An electromagnetic locking mechanism 410 is fixedly mounted on the upper outer wall of the hopper body 401. The electromagnetic locking mechanism 410 includes an electromagnetic drive coil and a telescopic locking tongue. The hopper cover 402 has a locking hole corresponding to the position of the electromagnetic locking mechanism 410. The electromagnetic locking mechanism 410 is electrically connected to the local signal junction box 440. When the electromagnetic locking mechanism 410 receives a first control signal from the edge computing gateway 200, the electromagnetic drive coil is energized, driving the telescopic locking tongue to extend in a straight line and insert into the locking hole of the hopper cover 402, mechanically locking the hopper cover 402 in the closed position. When a second control signal is received, the telescopic locking tongue retracts, releasing the mechanical lock on the hopper cover 402.

[0030] An open / closed state detection sensor 420 is fixedly mounted on the upper surface edge of the hopper body 401. The open / closed state detection sensor 420 is configured as a contact limit switch or a non-contact Hall sensor. The open / closed state detection sensor 420 is electrically connected to a local signal junction box 440. When the hopper cover 402 is in the closed position and is pressed down or close to the sensing end of the open / closed state detection sensor 420, the sensor generates a closed state electrical signal; when the hopper cover 402 leaves the closed position, the sensor generates an open state electrical signal. The closed and open state electrical signals are transmitted to the edge computing gateway 200 through the local signal junction box 440.

[0031] A near-field communication (NFC) reader 430 is embedded in the outer wall of the hopper body 401 on the side facing the operating channel. The NFC reader 430 is connected to a local signal junction box 440 via a shielded data cable. The NFC reader 430 is configured to continuously transmit a radio frequency query signal at a specific frequency. When a handheld smart terminal 500 enters the sensing range of the NFC reader 430 (0-10 cm), the NFC reader 430 reads the identification data fed back by the handheld smart terminal 500 and uploads this data to the edge computing gateway 200 via the local signal junction box 440.

[0032] The local signal junction box 440 is sealed and installed on the outer wall of the hopper body 401. The local signal junction box 440 integrates signal conditioning circuitry and terminal blocks. The control lines of the electromagnetic locking mechanism 410, the signal lines of the open / closed status detection sensor 420, and the data lines of the near-field communication reader 430 are all connected to the local signal junction box 440. The local signal junction box 440 is connected to the input / output interface of the edge computing gateway 200 via an industrial bus cable, enabling bidirectional data interaction between the intelligent hopper unit 400 and the control system.

[0033] See attached document Figure 1 The data acquisition link provided by this invention is configured to establish a deterministic data transmission channel between the injection molding machine unit 300, the intelligent hopper unit 400, and the central verification server 100. The data acquisition link mainly consists of an edge computing gateway 200 and its internally integrated protocol conversion module 210, data cleaning module 220, and uplink communication module 230.

[0034] The edge computing gateway 200 is physically connected to the controller 310 inside the injection molding machine unit 300 via hardwiring. The connection interface uses an industrial Ethernet interface or an RS-485 serial communication interface. The protocol conversion module 210 is configured to load a pre-installed industrial communication protocol driver, which is compatible with the brand and model of the controller 310. The protocol conversion module 210 sends read commands to the controller 310 at a preset sampling frequency of 10 Hz to 100 Hz, polling a specific register address within the controller 310.

[0035] The polled register address mappings include: the first address mapped to the screw's real-time position data, the second address mapped to the current injection pressure data, the third address mapped to the screw's rotational torque data, the fourth address mapped to the screw's current rotational speed data, and the fifth address mapped to the injection molding machine's operating cycle status word. The protocol conversion module 210 receives the raw hexadecimal data fed back by the controller 310 and converts it into a decimal value.

[0036] The data cleaning module 220 is configured to perform engineering dimension conversion and timestamp on the raw decimal values ​​obtained from the controller 310. For the digitization conversion of analog signals, the data cleaning module 220 uses a linear transformation formula to calculate the physical engineering values. The calculation formula is as follows: ; in: This represents the raw discrete digital value read from the register of controller 310; This represents the proportional gain coefficient of the corresponding sensor, which is determined by the ratio of the sensor's range to the analog-to-digital conversion resolution. This represents the zero-point drift correction constant or the base bias value; This indicates the converted engineering value with physical units, including millimeters, bar, newton-meter, or revolutions per minute.

[0037] The data cleaning module 220 also appends a unified system timestamp to each data packet it reads. This timestamp is synchronized with the central verification server 100 based on the network time protocol, with the synchronization accuracy controlled at the millisecond level.

[0038] The edge computing gateway 200 is also physically connected to the local signal junction box 440 of the smart hopper unit 400 via a digital input / output interface (Digital I / O). The edge computing gateway 200 sends a voltage signal to the local signal junction box 440 through its output port to drive the electromagnetic locking mechanism 410. The edge computing gateway 200 receives high and low level signals from the open / closed status detection sensor 420 transmitted from the local signal junction box 440 through its input port. The edge computing gateway 200 is connected to the local signal junction box 440 via a serial communication port to receive RFID tag data uploaded by the near-field communication reader 430.

[0039] The uplink communication module 230 is configured to establish a persistent connection with the central verification server 100 via the TCP / IP protocol stack. The uplink communication module 230 encapsulates the structured data object processed by the data cleaning module 220 into a data packet in JavaScript object abbreviation format. The uplink communication module 230 sends the encapsulated data packet to the message queue interface of the central verification server 100. The central verification server 100 parses the received data packet and stores it in a time-series database for subsequent verification algorithm calls.

[0040] See attached document Figure 2 In the material feeding permission control process based on spatiotemporal dual constraints provided by the present invention, the task binding and information flow verification sub-process is mainly executed collaboratively by the order management module 110 in the central verification server 100, the scanning verification module 510 in the handheld smart terminal 500, and the data interface module 520.

[0041] At the initial stage of the process, the operator activates the material feeding application on the handheld smart terminal 500. The handheld smart terminal 500 uses the optical lens of the scanning verification module 510 to acquire the image of the device QR code affixed to the control panel of the injection molding machine unit 300, and parses out the unique device identifier. The data interface module 520 will parse the obtained device identifier. and current operation timestamp It is encapsulated as a query request message and sent to the central verification server 100.

[0042] After receiving the query request message, the order management module 110 of the central verification server 100 performs a retrieval operation in the pre-stored production plan database. The retrieval criteria are set as follows: matching the equipment identifier field. The plan status field is set to "Executing". If the search result is empty, the order management module 110 returns an error response code indicating no production task; if the search is successful, the order management module 110 locks the retrieved production order record. .

[0043] Production order records This dataset contains the Bill of Materials (BOM) associated with the order. Bill of Materials Dataset This defines all the legal raw material information required for the current production task, and its data structure is represented as follows: ; in: Indicates the first Permitted raw material material codes; Indicates the first The standard consumption or total demand of a single mold for a certain type of raw material; This indicates the total number of raw material types included in the order.

[0044] The order management module 110 will store the bill of materials dataset. The included list of material numbers and basic order information are packaged and returned to the handheld smart terminal 500. The handheld smart terminal 500 renders the task interface to be fed into the machine on the display interaction module, which displays the current machine number, the product model being produced, and the specifications of the raw materials to be fed into the machine.

[0045] Subsequently, the operator used a handheld smart terminal 500 to scan the batch barcode on the outer packaging bag of the raw materials. The scanning verification module 510 decoded the batch barcode and extracted the material number of the raw material to be added. Batch number and the weight of a single package The handheld smart terminal 500 performs local logic comparison to determine the extracted material number. Does it exist in the list of material numbers obtained from the server? middle.

[0046] like The handheld smart terminal 500 determines that the material is mismatched, immediately triggers an audible and visual alarm, and prohibits the generation of material feeding requests. If... The handheld smart terminal 500 generates a pre-feeding request token. The pre-feeding request token Includes device identifier Raw material batch number and the amount of materials to be fed in this application. The status is marked as pending physical verification. At this time, the handheld smart terminal 500 prompts the operator to go to the smart hopper unit 400 to confirm the physical location.

[0047] See attached document Figure 2 The physical location and action interlocking mechanism provided by this invention is jointly executed by the location verification module 115 inside the central verification server 100 and the edge computing gateway 200 deployed on the production site. This mechanism aims to establish a unique correspondence between raw material input permissions and the spatial coordinates of physical equipment, preventing operators from performing material input actions on the wrong machine.

[0048] The location verification module 115 establishes an encrypted communication channel with the handheld smart terminal 500. When the operator arrives at the designated injection molding machine unit 300, the handheld smart terminal 500 reads the location identification tag fixedly installed on the outer wall of the smart hopper unit 400. This location identification tag stores a unique equipment physical address code. The handheld smart terminal 500 will read the... and the current scan timestamp Package and upload to the location verification module 115.

[0049] The location verification module 115 retrieves the current production work order data and extracts the preset target equipment code. The location verification module 115 will upload... and A binary bit comparison is performed. Only when the two are completely identical, the location verification module 115 generates an unlock authorization token with a time window restriction and sends it to the corresponding edge computing gateway 200.

[0050] The edge computing gateway 200 connects via hardwiring to the electromagnetic locking mechanism 410 and the open / closed status detection sensor 420 on the intelligent hopper unit 400. The edge computing gateway 200 has a built-in hardware interlock logic controller, which calculates the drive signal for the electromagnetic locking mechanism 410 based on the received authorization token and real-time clock signal. Drive signal The logical operation formulas are as follows: ; in: Indicates at time The control level output to the electromagnetic locking mechanism 410; This indicates a high level (unlock / magnetize). This indicates a low level (lock-in / demagnetization release). This indicates the physical address code of the device obtained from the actual on-site scan; This represents the unique and valid equipment code specified in the production work order; This indicates the moment when the server confirms a location match and generates a token; This parameter indicates the valid lifespan of the authorization token. It is set to 30 to 60 seconds and is used to limit the maximum allowable time interval between scan confirmation and the execution of the opening cover action.

[0051] When the drive signal for At this time, the electromagnetic locking mechanism 410 performs a mechanical unlocking action, allowing the hopper cover to be physically opened. Simultaneously, the edge computing gateway 200 polls the signal status of the opening / closing sensor 420 at a frequency of 100 milliseconds. If in Within the time window, the opening / closing state detection sensor 420 detects that the hopper cover changes from a closed state to an open state. The edge computing gateway 200 determines that a valid physical feeding action has begun and immediately sends an event message indicating the start of the action to the central verification server 100.

[0052] If time Exceeded The boundary was reached, and the open / closed state detection sensor 420 still did not detect an open signal, so the edge computing gateway 200 forcibly closed the boundary. Set as The electromagnetic locking mechanism 410 immediately relocks, and the corresponding authorization token becomes invalid. This logic ensures that feeding permissions are only valid when the human and machine physical locations coincide and within a tight time window, preventing operators from accidentally feeding materials to machine B after scanning the code on machine A.

[0053] After the physical feeding action is completed, when the open / closed state detection sensor 420 detects the hopper cover closing signal again, the edge computing gateway 200 drives the electromagnetic locking mechanism 410 to enter the locked state again and sends an action termination confirmation frame to the position verification module 115. The position verification module 115 then closes the current operation session until the next valid scan request is initiated.

[0054] See attached document Figure 2 The virtual session management method provided by this invention is executed by the session management module 120 integrated within the central verification server 100. This method aims to solve the batch consistency control problem when the same batch of raw materials needs to be delivered multiple times by different personnel in a continuous flow production mode.

[0055] In the initial stage of production task initiation, the session management module 120 is in a listening state. When the handheld smart terminal 500 initiates a material feeding request for a specific injection molding machine unit 300, the session management module 120 first queries whether there is already an active material feeding session for that injection molding machine unit 300.

[0056] If the query results indicate that there are no active sessions currently, and the data delivery request passes the aforementioned BOM matching verification, the session management module 120 determines that this data delivery is the first data delivery. The session management module 120 then instantiates a new session object in the in-memory database. The session object Defined as a data structure vector containing multidimensional state attributes, its mathematical expression is as follows: ; in: This represents a globally unique session sequence number generated by the system. This indicates the physical number of the injection molding machine unit bound to this session; This indicates the anchor batch number, which is the unique raw material batch code that is only allowed to be added during this session. Its value is initialized to the batch number obtained from the first package of material added. This represents the server's standard timestamp at the time the session was created; This represents the session's lifecycle status word, with an initial value of ACTIVE. This represents the set of operator IDs involved in the session, used for audit trails.

[0057] Once the session object Created and When set to ACTIVE, the session management module 120 applies a single batch locking strategy to the injection molding machine unit 300 at the logical level. In this state, for any new material feeding request to the injection molding machine unit 300, the session management module 120 no longer performs basic BOM list matching, but instead performs strict session anchoring comparison.

[0058] Assume that a new scanning batch number for this machine is subsequently received from any handheld smart terminal 500. The session management module 120 executes the following discrimination function. : ; when The output is At that time, the session management module 120 determines that the request is a legitimate renewal material delivery. The system allows the unlocking process to continue and appends the current operator ID to the set. Simultaneously, the last activity time of the session is updated. This logic ensures that even in shift changes or multi-person collaboration scenarios, as long as the previous batch is not completed, any session belonging to the BOM list but with a different batch number will be updated. All raw materials will be intercepted by the system, thus physically eliminating the possibility of batch mixing.

[0059] when The output is At that moment, the session management module 120 determined it to be an attempt at batch cross-contamination. The system immediately rejected the feeding request and returned a message to the handheld smart terminal 500 indicating that the batch was locked: Please feed the batch. Alternatively, you can first end the blocking command for the current session.

[0060] The session management module 120 is also configured with a status transition interface. The session management module 120 will only initiate a status transition when it receives an explicit clearing / batch completion instruction, or when the inventory consumption logic in subsequent chapters determines that the hopper is empty. Updated to CLOSED. After the session is closed, the batch lock of the injection molding unit 300 is released, allowing the next different batch of raw materials to trigger a new first-pack feeding logic.

[0061] See attached document Figure 2 The cross-process physical interlocking mechanism provided by this invention is mainly achieved through the collaboration of the interlocking control module 130 deployed inside the central verification server 100 and the downstream execution unit 600 located downstream of the injection molding machine unit 300. The downstream execution unit 600 includes a product removal robot 610, a conveyor belt control box 620, and a finished product label printer 630.

[0062] The interlock control module 130 establishes a deterministic communication connection with the downstream execution unit 600 via an industrial fieldbus or industrial Ethernet. The interlock control module 130 periodically reads the session status attributes from the aforementioned session management module 120. and anchor batch number The interlock control module 130 is configured to calculate the enable control signal for the downstream execution unit 600 in real time based on the current raw material feeding status. .

[0063] Enable control signal The logical operation formulas are as follows: ; in: This indicates the current virtual session status bound to this injection molding machine unit 300; This indicates that the current session is in a legitimate, locked single-batch production session; This indicates a global system anomaly flag. It is triggered when the hopper level is zero or the rheological fingerprint verification fails. Set as Under normal circumstances, it is .

[0064] When the calculation result for At this time, the interlock control module 130 continuously sends a high-level heartbeat signal or a logic truth value instruction to the downstream execution unit 600. In this state, the product removal robot 610 executes the standard stacking procedure, the conveyor belt control box 620 drives the conveyor belt to the qualified product area, and the finished product label printer 630 allows printing of labels containing the current anchored batch number. Product traceability labels.

[0065] When the calculation result jump to When this occurs, it indicates that the current feeding end is in a state of undefined batch, session ended, or there is a risk of raw material abnormality. The interlock control module 130 immediately cuts off the permission signal sent to the downstream execution unit 600, triggering a physical blocking action. Specifically, after receiving the blocking signal, the product removal robot 610 executes a forced diversion path, placing the injection molded parts ejected from the mold of the injection molding machine unit 300 into the isolation area or waste chute, physically preventing the injection molded parts produced during this period from entering the qualified product logistics line.

[0066] Simultaneously, the conveyor belt control box 620 responds to the blocking signal, immediately stopping the drive motor or triggering the diversion baffle to prevent unidentified batches of products in transit from flowing to the next process. The finished product label printer 630 enters a locked state, refusing to execute any printing tasks and clearing any label data remaining in the print buffer to prevent incorrect labeling where the actual product does not match the label batch.

[0067] In addition, the interlock control module 130 also maintains a batch switching buffer counter. When the session state When changing from ACTIVE to CLOSED, the interlock control module 130 does not immediately release the lock on the downstream execution unit 600, but waits for a new session to be established. Once the new session is established and a new batch of raw materials is fed in, the interlock control module 130 calculates the number of transition cycles that need to be emptied based on the theoretical volume of the injection molding machine screw and the consumption per cycle. In the subsequent Within each production cycle, even though a new session has been activated, the interlock control module 130 forcibly... Keep as Continue this process until the transition material is completely discharged, ensuring strict alignment of the physical product with the new batch data.

[0068] This invention provides a single-mold raw material consumption calculation model, which runs on an algorithm engine module 140 integrated within a central verification server 100. The algorithm engine module 140 is configured to establish a deterministic mapping relationship from mechanical stroke to raw material quality based on the physical motion parameters of the injection molding machine unit 300.

[0069] The algorithm engine module 140 first establishes a data channel with the injection molding machine controller through the edge computing gateway 200, collecting stroke data and process status data for each injection cycle in real time. Stroke data includes the metering start position (i.e., the end position of material storage) and the injection termination position (i.e., the position of residual material pad). Process status data includes the current barrel temperature and melt back pressure. The algorithm engine module 140 internally stores a hardware parameter table corresponding to the injection molding machine unit 300, which includes the screw diameter, the effective length of the screw metering section, and the check ring volumetric efficiency coefficient.

[0070] The algorithm engine module 140 performs a single-mold melt mass integral calculation based on the collected real-time data and pre-stored hardware parameters. This calculation aims to eliminate interference from environmental factors and accurately quantify the actual raw material reduction during a single injection molding process. The algorithm engine module 140 uses the following core mass calculation formula to determine the raw material consumption mass per mold. : ; in: This indicates the actual mass of molten material injected into the mold cavity during the current injection cycle, expressed in grams. This indicates the physical outer diameter of the injection molding machine screw, in centimeters. This indicates the coordinates of the metering start position before the screw performs the injection action, in centimeters. This indicates the final position coordinates of the screw after completing the injection and pressure holding actions, in centimeters. Indicates the current barrel temperature and injection pressure The raw material melt density under the given conditions is obtained by interpolation from the internal PVT (pressure-specific volume-temperature) characteristic database called by the algorithm engine module 140, and the unit is grams per cubic centimeter. This represents the volumetric efficiency correction factor of the injection system, used to compensate for volume loss caused by melt backflow and compression escape. This factor is a dimensionless constant.

[0071] The algorithm engine module 140 triggers the above calculation logic at the moment the injection cycle ends, and calculates the result. The values ​​are written to the accumulator register in dynamic memory. For minute fluctuations during continuous production, the algorithm engine module 140 is configured to handle several consecutive cycles. The numerical values ​​are subjected to an arithmetic mean to eliminate the impact of signal noise on measurement accuracy.

[0072] The algorithm engine module 140 will also calculate the raw material consumption mass per single-module. The weight is compared in real time with the theoretical unit weight in the standard process card. When the absolute value of the deviation exceeds the preset tolerance range, the algorithm engine module 140 generates a measurement anomaly flag and associates the flag with the current production batch record for subsequent quality traceability and interception analysis.

[0073] See attached document Figure 2 The real-time inventory hedging and anomaly detection mechanism provided by this invention is executed by the algorithm engine module 140 inside the central verification server 100. This mechanism is based on a single-module consumption calculation model and aims to achieve closed-loop monitoring of the integrity of material flow by comparing virtual inventory calculations in the digital world with sensor status feedback in the physical world.

[0074] Algorithm engine module 140 first initializes a virtual inventory register in memory. When the handheld smart terminal 500 described in Embodiment 2 completes the legitimate material feeding confirmation operation and uploads the material feeding request token... At that time, the algorithm engine module 140 parses the raw material net weight data contained in the token and adds the value to the virtual inventory register. This process establishes the initial boundary conditions for the dynamic equilibrium calculation.

[0075] As injection molding unit 300 begins continuous production, algorithm engine module 140 operates in each injection molding cycle. At the end, the raw material consumption mass of the single-module cycle is calculated using the aforementioned single-module consumption calculation model. The algorithm engine module 140 performs real-time deduction calculations and updates the virtual inventory register. The value. The number. Real-time inventory levels at the end of each cycle The iterative calculation formula is as follows: ; in: Indicates the first The theoretical remaining inventory mass after the end of each cycle, in grams; Indicates the first The theoretical remaining inventory mass at the end of each cycle, in grams; Indicates the first The actual mass of raw materials consumed in each cycle, expressed in grams.

[0076] While performing the deduction calculation, the algorithm engine module 140 continuously monitors the physical material shortage signal emitted by the injection molding machine unit 300 or the smart hopper unit 400 through the edge computing gateway 200. This physical shortage signal This originates from a capacitive proximity switch installed in the neck of the hopper or a plasticizing timeout alarm bit inside the injection molding machine controller. When detected... When the signal transitions to an active state (logic 1), or when the calculated real-time inventory level... When the zero point is reached, the algorithm engine module 140 triggers the exception judgment logic.

[0077] Algorithm engine module 140 calculates the mass balance state word based on the current physical and computational states. The decision logic is implemented using the following piecewise functions: ; in: Indicates the status of the physical material shortage sensor; This indicates that the space is physically empty; This means that there are still materials physically present; This indicates the current theoretical remaining inventory calculated by the system. This indicates the system's allowed cumulative metering error threshold, which is set to 0.5% to 1.0% of the total feed amount. This indicates an abnormal material loss status, meaning that the physical hopper is empty but the system calculates that there is still a large inventory. This indicates that there is raw material leakage, theft, or severe wear of the metering screw, resulting in the actual amount of material discharged in a single instance being greater than the theoretical calculation value. This indicates that no abnormal feeding status was recorded, meaning that the system calculated that the inventory was exhausted (or even negative), but there was still material in the physical hopper and the machine continued to produce, indicating that there was illegal feeding behavior that bypassed the scanning process or that foreign objects were mixed in. This indicates a production status that is within the normal tolerance range.

[0078] when Determined as or Upon this event, the algorithm engine module 140 immediately generates a high-level alarm event and sends a shutdown interlock command to the injection molding machine unit 300. Simultaneously, the system automatically freezes the quality traceability records for the current production batch, marking the batch as having failed quality balance verification, until management personnel intervene to investigate and unlock the system.

[0079] See attached document Figure 2 The physical rollback logic for the cancellation operation provided by this invention is executed by the transaction rollback module 150 integrated within the central verification server 100. This mechanism is configured to handle cancellation requests issued by operators after performing material feeding scanning, and to determine the rollback strategy for logical data based on the actual state of the physical world.

[0080] The process begins with the handheld smart terminal 500 sending a reversal request message containing the target device identifier and operation sequence number to the central verification server 100. Upon receiving the request message, the transaction rollback module 150 immediately locks the associated virtual session record and initiates a physical status query command to the edge computing gateway 200. The edge computing gateway 200 retrieves the timestamped sensor event logs stored locally. The edge computing gateway 200 extracts data from the material feeding scan time... Until the time of withdrawal request The signal waveform of the sensor 420 for detecting the opening and closing status within this time window.

[0081] The transaction rollback module 150 performs a physical intervention determination based on the sensor signal waveform. The determination logic is based on the physical action characteristic function defined below. : ; in: It represents any point in time within the time window; Indicates at time The physical open / closed state of the hopper cover; This indicates a state of no physical intervention, meaning the hopper cover remains closed throughout the entire operating window. This indicates that physical intervention has occurred, meaning that the hopper cover has been opened at least once during the operating window.

[0082] when The calculation result is At this point, the transaction rollback module 150 determines that the revocation request is a erroneous operation without physical consequences. The transaction rollback module 150 performs a purely logical rollback operation: directly deletes the virtual inventory record that is about to take effect, sends an instruction to the edge computing gateway 200 to cancel the pre-authorization status of the electromagnetic locking mechanism 410, and returns a confirmation message of successful revocation to the handheld smart terminal 500.

[0083] when The calculation result is At this point, the transaction rollback module 150 determines that the operation has produced physical consequences, namely, the raw material has been actually fed into the hopper. The transaction rollback module 150 then refuses to directly cancel the operation and forces the injection molding machine unit 300 into a cleaning lock mode. In cleaning lock mode, the injection molding machine unit 300 is prohibited from performing normal production molding cycles and is only allowed to perform air injection and material discharge operations.

[0084] Transaction rollback module 150 based on the quality of the mistakenly added raw materials And the physical retention volume of the injection molding machine barrel, calculate the minimum evacuation and cleaning quality that must be performed. The calculation formula is as follows: ; in: This indicates the total mass of melt that needs to be discharged to unlock, expressed in grams. This indicates the net weight of the mistakenly added raw materials for which this application for withdrawal has been submitted, in grams. This indicates the total geometric volume of the front section and metering section of the injection molding machine's barrel screw, expressed in cubic centimeters. This indicates the melt density at the current temperature, expressed in grams per cubic centimeter. This represents the safety cleaning factor, ranging from 1.5 to 2.0, used to ensure that residual materials are completely replaced.

[0085] During the air-jet discharge operation in the injection molding unit 300, the algorithm engine module 140 continuously accumulates the discharge volume calculated by the screw metering integral. Only when the accumulated discharge volume reaches or exceeds... When the transaction rollback module 150 releases the cleaning lock mode, it allows the system to be restored to the initial state before the rollback, and generates an audit log containing records of the physical cleaning execution.

[0086] In the closed-loop material property secondary verification based on rheological fingerprinting provided by this invention, the determination of the lag period and sampling window is executed by the rheological analysis module 160 integrated within the central verification server 100. This process aims to solve the time / module delay problem between the physical feeding action and the actual melt property changes at the injection molding machine nozzle, ensuring that the system captures data features only within the stable range after the raw material has been completely replaced.

[0087] Upon receiving the material feeding completion confirmation signal, the rheology analysis module 160 immediately reads the current inventory status and equipment operating parameters. The rheology analysis module 160 identifies a physical volumetric stagnation area between the hopper neck and the injection molding machine nozzle. This stagnation area includes the remaining old material at the bottom of the hopper, the solid conveying zone at the rear of the barrel, the compression and melting zone in the middle section, and the metering and storage zone at the front. To accurately pinpoint the moment when newly fed material reaches the nozzle and forms a stable melt flow, the rheology analysis module 160 executes a lag cycle. The calculation.

[0088] The rheological analysis module 160 retrieves the theoretical remaining inventory of the hopper calculated by the algorithm engine module 140. Single-mold raw material consumption quality Based on the device hardware parameters, the number of hysteresis cycles to be skipped is calculated using the following formula. : ; in: This represents the number of production cycles required from the completion of the feeding process to the emergence of the rheological properties of the new material, rounded up to the nearest integer. This indicates the mass of old material in the hopper that has not yet entered the machine barrel at the moment of feeding, expressed in grams. This represents the geometric free volume of the injection molding machine screw throughout its entire stroke, expressed in cubic centimeters. This indicates the melt density at the current process temperature, expressed in grams per cubic centimeter. This represents the average filling factor of the screw groove depth. This factor is determined by the screw design parameters and the efficiency of the feeding section, and its value is usually between 0.8 and 0.9. This indicates the actual injection weight per mold under the current production process, in grams; This represents the axial mixing dispersion coefficient, used to correct for the melt mixing effect caused by non-plunging flow. This coefficient ensures that the additional cycle required to completely flush out the overlying material is calculated, and its value is always greater than 1.0.

[0089] The rheological analysis module 160 calculates the results... Set a silent counter. In the subsequent... During each injection molding cycle, although the rheological analysis module 160 continuously receives sensor data uploaded by the injection molding machine unit 300, it does not perform feature comparison calculations, but only performs data caching or discarding to eliminate unstable data interference during the mixed transition period.

[0090] When the production cycle count reaches At this time, the rheological analysis module 160 automatically activates the effective sampling window. This window defines a continuous monitoring interval used to capture the steady-state rheological fingerprint of newly input raw materials. The effective sampling window terminates at a certain time. Determined by the following logic: ; in: This indicates the end cycle number of this material feeding verification sampling; This represents the preset minimum sample size constant, used to ensure statistical significance. This value is set to the number of cycles of continuous and stable production (e.g., 20 cycles).

[0091] From arrive During the sampling window, the rheological analysis module 160 frequently acquires the melt pressure curve of the injection molding machine unit 300 during the injection stage. Screw position curve The rheological analysis module 160 maps these raw waveform data into viscosity index and flow resistance characteristic values, and stores them in a temporary verification buffer, awaiting consistency comparison with the standard rheological fingerprint model. Only when the mean value of the characteristic data within the sampling window converges within the preset tolerance range will the system finally determine that the physical properties of the batch of raw materials are qualified.

[0092] The rheological feature extraction method provided by this invention is executed by the rheological analysis module 160 integrated within the central verification server 100. This method is configured to convert high-frequency discrete sensor data uploaded by the injection molding machine unit 300 into quantitative indicators characterizing the physical flow properties of the raw material, i.e., rheological fingerprints.

[0093] The rheology analysis module 160 establishes a synchronous data acquisition channel with the controller of the injection molding machine unit 300 via an industrial Ethernet interface. During each injection cycle within the effective sampling window, the rheology analysis module 160 acquires the current real-time injection pressure in parallel at sampling intervals ranging from 10 milliseconds to 50 milliseconds. and real-time screw forward speed The rheological analysis module 160 first performs interval truncation on the acquired raw time series data. Based on the data from the screw position sensor, the rheological analysis module 160 retains only the data segments where the screw stroke is within the range of 10% to 90% of the total injection stroke, and removes the inertial acceleration zone at the start of injection and the pressure holding switching zone at the end of injection, ensuring that the data used for calculation originates from the steady-state laminar flow stage of the melt.

[0094] The rheological analysis module 160 performs numerical integration on the truncated valid data segment to calculate the apparent viscosity index for the current period. The rheological analysis module 160 uses the following rheological feature extraction formula: ; in: This represents the apparent viscosity index of the raw material calculated during the current injection molding cycle, in Pascals per second. This represents the total number of discrete sampling points collected within the valid data segment; Indicates the first The actual injection pressure value recorded at each sampling point at any given time is in bar. Indicates the first The actual screw forward speed value recorded at each sampling point is in millimeters per second. This indicates the depth of the screw groove in the metering section of the injection molding machine screw, in millimeters. This indicates the effective length of the screw metering section, in millimeters. This represents the shear thinning calibration coefficient, a dimensionless constant used to correct viscosity deviations at different shear rates based on the power-law model of non-Newtonian fluids. Its value is determined by the rheological test curve of the reference material under standard processes.

[0095] The rheological analysis module 160 completes a single cycle. After calculation, the result is stored in a temporary feature register. Once all periods within the effective sampling window (e.g., 20 modules) have been calculated, the rheological analysis module 160 performs statistical convergence calculations. The rheological analysis module 160 calculates this set of... The arithmetic mean and standard deviation of the values. Only when the standard deviation is less than a preset dispersion threshold, the rheological analysis module 160 solidifies the arithmetic mean as the final rheological fingerprint of the current production batch. .

[0096] The rheological analysis module 160 will then calculate the... The standard fingerprint template of this raw material grade A comparison is performed. If the absolute value of the deviation between the two exceeds the allowable tolerance range (e.g., ±3%), the rheological analysis module 160 determines that the physical properties of the currently input raw material do not match the raw material specified in the BOM, and then triggers a rheological verification failure alarm, and sends a shutdown blocking command to the interlock control module 130 to prevent the potential risk of incorrect materials from continuing to the subsequent production process.

[0097] The fingerprint comparison and closed-loop control process provided by this invention is executed by the rheological analysis module 160 and the interlocking control module 130 within the central verification server 100. This process aims to automatically determine the physical consistency of the current raw material based on the rheological feature values ​​extracted in the aforementioned steps, and directly intervene in the operating status of the injection molding machine unit 300 based on the determination result, thus forming a quality closed loop.

[0098] The rheological analysis module 160 first retrieves the standard raw material rheological fingerprint template specified for the current production task from the production work order database. This template is based on a baseline viscosity index pre-measured and stored using standard raw materials under standard process conditions. The rheological analysis module 160 receives the rheological fingerprint of the current batch calculated by a feature extraction algorithm. .

[0099] The rheological analysis module 160 performs relative deviation calculations to quantify the differences in physical properties between the current production raw materials and standard raw materials. Relative deviation rate. The calculation formula is as follows: ; in: The relative deviation rate of the rheological fingerprint is expressed as a percentage. This represents the measured rheological fingerprint value extracted within the current valid sampling window; This represents the standard rheological fingerprint value of the raw material of this grade that is preset in the database.

[0100] Obtain the relative deviation rate Then, the rheological analysis module 160 performs the analysis based on the preset quality control threshold. Execution logic determination. This is the upper limit of the allowable fluctuation range set according to the molding accuracy requirements of the final product, typically set to 3.0% to 5.0%. The rheological analysis module 160 generates control action commands based on the judgment results. Its logical function is defined as follows: ; in: This indicates a release order, meaning that the rheological properties of the raw material meet the process requirements; A blocking command indicates a significant shift in the rheological properties of the raw material, suggesting a risk of incorrect material selection or degradation.

[0101] when Determined as At that time, the rheological analysis module 160 will display the current... Numerical values ​​and The deviation rate is written into the electronic record of the production batch as a certificate of conformity for quality traceability. The system maintains the normal authorization status of injection molding machine unit 300, allowing it to continue executing subsequent continuous production cycles.

[0102] when Determined as At that time, the rheological analysis module 160 immediately sends a high-priority trigger signal to the interlock control module 130. The interlock control module 130 then sends a cycle prohibition command to the controller of the injection molding machine unit 300 through the edge computing gateway 200. This command forces the injection molding machine unit 300 to prevent the start of the next injection molding cycle after completing the current molding cycle, keeping the equipment in a standby state.

[0103] Simultaneously, the interlock control module 130 will generate an alarm event indicating a rheological verification failure on the central management terminal and illuminate the on-site audible and visual alarm tower. At this time, the production batch record is automatically marked as frozen until an authorized process engineer empties the abnormal raw materials from the equipment and manually reviews and unlocks the alarm, at which point the system can restore production access.

[0104] See attached document Figure 2 The server-side software functional module deployment architecture provided by this invention runs within the computing unit of the central verification server 100. This architecture adopts a modular and layered design, realizing data interaction and logical collaboration between various functional units through an internal message bus. The central verification server 100 is internally configured with a non-volatile storage medium that stores computer program instructions that can be executed by a processor. These instructions are instantiated into multiple independent logical functional modules during runtime.

[0105] The logical function modules include a communication interface module 102, a location verification module 115, an interlock control module 130, an algorithm engine module 140, a transaction rollback module 150, a rheological analysis module 160, and a data persistence module 170. Data is transferred between modules through well-defined API interfaces or shared memory areas, ensuring decoupling of business logic and parallel processing capabilities.

[0106] The communication interface module 102 is located at the bottom layer of the system architecture and is configured to manage all physical connections and protocol conversions between the central verification server 100 and external hardware devices. The communication interface module 102 establishes and maintains TCP / IP network connections with the edge computing gateway 200 and the handheld smart terminal 500. The communication interface module 102 integrates a multi-protocol parser, used to decapsulate MQTT messages or Modbus-TCP packets from the edge into standardized JSON format internal data frames, and to encapsulate control commands issued by the server into communication messages recognizable by the corresponding devices.

[0107] The location verification module 115 is connected to the communication interface module 102 and is configured to handle identity and location authentication services during the material feeding request phase. The location verification module 115 receives scan data containing the device's physical address code and operation timestamp, and compares it with production order data stored in the data persistence module 170. The output of the location verification module 115 is connected to the input of the interlock control module 130, and sends a pre-authorization signal allowing unlocking only to the interlock control module 130 when verification is successful.

[0108] The interlock control module 130, acting as the system's decision-making and execution center, is configured to integrate the status inputs from various business modules and generate the final physical equipment control commands. The interlock control module 130 receives pre-authorization signals from the position verification module 115, inventory status signals from the algorithm engine module 140, and quality judgment signals from the rheological analysis module 160. The interlock control module 130 performs logical AND operations and priority arbitration, and issues electromagnetic lock drive commands to the edge computing gateway 200 or stop interlock commands to the injection molding machine unit 300 via the communication interface module 102.

[0109] The algorithm engine module 140 is configured to perform core numerical calculations for raw material consumption and inventory balance. The algorithm engine module 140 subscribes to the real-time process parameter stream of the injection molding machine published by the communication interface module 102, including screw position, injection pressure, and melt temperature. Based on the single-mold consumption calculation model and real-time inventory hedging logic described in the preceding embodiments, the algorithm engine module 140 calculates the real-time remaining inventory and quality balance status, and writes the calculation results to the data persistence module 170 for storage in real time. Simultaneously, it sends abnormal status words to the interlock control module 130.

[0110] The transaction rollback module 150 is configured to handle undo operation requests and recovery procedures after exceptions. In response to an undo command, the transaction rollback module 150 retrieves sensor logs within a specified time window via the communication interface module 102. The transaction rollback module 150 determines whether to execute the cleaning logic based on the physical action characteristic function, and requests the interlock control module 130 to lock the injection molding machine if necessary, until the cleaning action meets the calculated minimum evacuation quality requirements.

[0111] The rheology analysis module 160 is configured to perform secondary verification based on the physical properties of the melt. Within the effective sampling window determined by the algorithm engine module 140, the rheology analysis module 160 acquires the pressure and velocity waveforms of the injection process at high frequency through the communication interface module 102. The rheology analysis module 160 calculates the rheological fingerprint feature value and compares it with the standard fingerprint template stored in the data persistence module 170. The release or blocking command generated by the comparison result is directly transmitted to the interlock control module 130 for closed-loop control of production permissions.

[0112] The data persistence module 170 is configured to manage the system's static basic data and dynamic business data. Static basic data includes the Bill of Materials (BOM), equipment hardware parameter tables, standard rheological fingerprint templates, and user permission tables. Dynamic business data includes production batch records, material feeding logs, real-time inventory snapshots, and alarm audit trail records. The data persistence module 170 provides a unified read / write access interface for all the above functional modules, ensuring data consistency and traceability.

[0113] See attached document Figure 1 The edge computing gateway 200 and handheld smart terminal 500 deployed in the industrial site provide the edge and terminal functions provided by this invention. This part of the architecture is configured to realize high-frequency acquisition of physical signals, local protocol conversion, execution of equipment actions, and presentation of human-machine interface.

[0114] The edge computing gateway 200 is physically connected to the electrical interfaces of the injection molding machine unit 300 control cabinet and the intelligent hopper unit 400. The edge computing gateway 200 includes multiple industrial communication interfaces. An RS-485 serial interface is connected to the PLC controller of the injection molding machine unit 300. Digital input / output interfaces are connected to the electromagnetic locking mechanism 410 and the open / closed status detection sensor 420. The edge computing gateway 200 internally runs a real-time embedded operating system configured to perform low-level data polling and preprocessing tasks.

[0115] The edge computing gateway 200 periodically polls the register address of the injection molding machine unit 300 via the Industrial Fieldbus protocol. The polling period is set to 10 milliseconds to 50 milliseconds. The edge computing gateway 200 acquires real-time screw position, injection pressure, and screw speed data. The edge computing gateway 200 performs a moving average filtering algorithm locally on the raw analog signal to eliminate signal glitches caused by electromagnetic interference in the field. The edge computing gateway 200 uses the following discrete filtering formula to calculate the effective sampled values ​​uploaded to the server. : ; in: Indicates the first The filtered physical quantity values ​​calculated at each sampling time; This parameter represents the length of the sliding window. It is set to an integer and is used to define the number of historical sampling points involved in the average calculation. Indicates the first The raw sensor values ​​collected at each sampling time.

[0116] The edge computing gateway 200 will calculate the results. The message frames are packaged into JSON format. The edge computing gateway 200 uploads the message frames to the central verification server 100 via TCP / IP protocol. The edge computing gateway 200 listens for control commands from the central verification server 100. When it receives an unlock command for the electromagnetic locking mechanism 410, the edge computing gateway 200 sends a high-level holding signal to the relay drive circuit through its digital output port. The edge computing gateway 200 has a built-in hardware watchdog timer. Once the edge computing gateway 200 detects that the communication link interruption with the central verification server 100 has exceeded a preset safety threshold, the edge computing gateway 200 automatically forces a reset of the digital output port to a low level. This action restores the electromagnetic locking mechanism 410 to its locked state.

[0117] The handheld smart terminal 500 is configured as a mobile front-end for operators to interact with the system. It integrates an optical barcode scanning module, an RFID reader / writer, an LCD screen, and a wireless network transceiver unit. The handheld smart terminal 500 runs a client application. This client application guides operators through standardized material feeding procedures.

[0118] The handheld smart terminal 500, triggered by the physical scan button, activates its optical scanning module to read the barcode on the raw material packaging bag and the hopper location label. The handheld smart terminal 500 performs local format verification on the read raw coded data. It discards damaged or invalid data that does not conform to the coding rules. Upon successful verification, the handheld smart terminal 500 generates an operation request message. This message includes the device's unique code, the material's unique code, and the terminal's local timestamp. The handheld smart terminal 500 then sends the operation request message to the central verification server 100.

[0119] The handheld smart terminal 500 receives feedback instructions from the central verification server 100 and drives the local human-machine interface accordingly. When a material feeding permission instruction is received, the handheld smart terminal 500 displays a green pass icon on the LCD screen. The handheld smart terminal 500 drives the built-in vibration motor to perform short pulse vibrations. When a material feeding prohibition instruction is received, the handheld smart terminal 500 displays a red warning icon and the specific error code. The handheld smart terminal 500 drives a buzzer to emit intermittent alarm sounds until the operator confirms the alarm information.

Claims

1. An injection material feeding mistake proofing and checking system for a safety system airbag, characterized by, The application relates to a production order verification system for injection molding machine, comprising: a central verification server storing a production order database and a bill of materials database, which is communicatively connected with an edge computing gateway and a handheld intelligent terminal; the handheld intelligent terminal is used for identifying the equipment identification of an injection molding machine unit and the raw material information of a raw material packaging bag, and the handheld intelligent terminal is used for sending the equipment identification and the raw material information to the central verification server; the central verification server is used for retrieving the production order database and the bill of materials database according to the equipment identification and the raw material information, and the central verification server is used for generating an authorization instruction after verification and sending the authorization instruction to the edge computing gateway; a smart hopper unit is installed at the feeding port of the injection molding machine unit, and the smart hopper unit comprises an electromagnetic locking mechanism; the edge computing gateway is physically connected to the controller of the injection molding machine unit, the edge computing gateway is connected to the smart hopper unit, and the edge computing gateway is used for driving the electromagnetic locking mechanism to perform an unlocking action in response to the authorization instruction.

2. The injection material feeding mistake proofing and checking system for a safety system airbag according to claim 1, characterized by, The smart hopper unit further comprises: a hopper body for temporarily storing raw materials; a hopper cover rotatably connected to the hopper body; an opening and closing state detection sensor installed at the edge of the hopper body and used for detecting the physical opening and closing states of the hopper cover; a near field communication reader / writer installed on the outer wall of the hopper body and used for communicating with the handheld intelligent terminal and confirming the physical position; a local signal junction box connected with the electromagnetic locking mechanism, the opening and closing state detection sensor and the near field communication reader / writer and connected to the input / output interface of the edge computing gateway through a signal cable, and the electromagnetic locking mechanism is fixed to the hopper body and used for locking and releasing the hopper cover.

3. The injection material feeding mistake proofing and checking system for a safety system airbag according to claim 1, characterized by, The edge computing gateway internally integrates: a protocol conversion module used for loading a driver matched with the controller and polling the register address of the controller; the register address mapping data comprises real-time screw position data, glue injection pressure data, screw rotation torque data and current screw rotation speed data; a data cleaning module used for converting original discrete digital data obtained from the controller into engineering values with physical units by using the proportional gain coefficient and zero point drift correction constant of the sensor and adding a time stamp synchronized with the central verification server to the engineering values.

4. The injection material feeding mistake proofing and checking system for a safety system airbag according to claim 2, characterized by, The central verification server internally integrates a position verification module, which performs the following operations: receiving the equipment physical address code and scanning time stamp sent by the handheld intelligent terminal after reading the position identification tag of the near field communication reader / writer; comparing the equipment physical address code with the target equipment code specified in the production work order, and generating an unlocking authorization token with a time window limit and issuing the unlocking authorization token to the edge computing gateway when the comparison is consistent, and the edge computing gateway drives the electromagnetic locking mechanism to perform an unlocking action only within the valid survival period of the unlocking authorization token; if the opening signal is not detected by the opening and closing state detection sensor within the valid survival period, the electromagnetic locking mechanism is forced to be relocked.

5. The injection material feeding mistake proofing and checking system for a safety system airbag according to claim 1, characterized by, The central verification server internally integrates a session management module to perform the following operations: A session object containing an anchor batch number is created when the first package feeding request passes the verification, and the state of the injection molding machine unit is marked as a single batch locking state; In the single batch locking state, anchor comparison logic is performed for subsequent feeding requests of the injection molding machine unit: If the subsequent scanned raw material batch number is consistent with the anchor batch number, it is determined to be a legal package feeding; If the subsequent scanned raw material batch number is inconsistent with the anchor batch number, it is determined to be batch cross-contamination and feeding is blocked.

6. The injection material feeding mistake proofing and checking system for a safety system airbag according to claim 3, characterized by, The central verification server internally integrates an algorithm engine module to perform the following operations: According to the screw diameter, effective length of the screw metering section, and collected metering start position data and shot termination position data of the injection molding machine unit, combined with melt density and volume efficiency correction coefficient, the single-mold raw material consumption mass is calculated; Based on the raw material feeding amount and the single-mold raw material consumption mass, a virtual inventory register is maintained to perform real-time deduction operations to obtain the theoretical remaining inventory; The theoretical remaining inventory is compared with the physical material shortage signal from at least one of the injection molding machine unit and the intelligent hopper unit, and based on the comparison result, it is determined whether there is material loss abnormality and unrecorded feeding abnormality.

7. The injection material feeding mistake proofing and checking system for a safety system airbag according to claim 6, characterized by, The central verification server internally integrates a rheological analysis module to perform the following operations: After the feeding action is completed, the number of lagging mold cycles required for the new material to reach the nozzle is calculated based on the theoretical remaining inventory and the single-mold raw material consumption mass, and a silent counter that needs to be skipped is set; After the silent counter ends, an effective sampling window is activated; In the effective sampling window, the real-time injection pressure and real-time screw advancing speed of the injection molding machine unit during the shot stage are collected, and based on the screw groove depth and effective length of the screw metering section, the apparent viscosity index of the raw material is calculated as a rheological fingerprint.

8. The injection material feeding mistake proofing and checking system for a safety system airbag according to claim 7, characterized by, The rheological analysis module is also used to compare the rheological fingerprint with a pre-stored standard rheological fingerprint template and calculate the relative deviation rate; When the relative deviation rate exceeds the pre-set quality control threshold, the central verification server internally integrates an interlock control module to determine that the rheological verification fails, send a periodic prohibition instruction to the injection molding machine unit, force the injection molding machine unit to stop after completing the current molding cycle, and freeze the production batch record.

9. The injection material feeding mistake proofing and checking system for a safety system airbag according to claim 2, characterized by, The central verification server internally integrates a transaction rollback module to perform the following operations: In response to a cancel feeding request, the signal waveform of the opening and closing state detection sensor stored by the edge computing gateway is retrieved; It is judged whether the hopper cover has been physically opened during the operation window period: If not, perform data rollback at the logical level; If it has been opened, force the injection molding machine unit into a cleaning locking mode until the cumulative discharge amount of the executed air shot discharge action reaches the minimum emptying and cleaning mass calculated based on the misfed raw material mass and barrel retention volume.

10. The injection material feeding mistake proofing and checking system for a safety system airbag according to claim 5, wherein The system also includes a downstream execution unit located downstream of the injection molding machine unit, which includes a product taking robot, a conveyor control box, and a finished product label printer; The central verification server integrates an interlock control module, which is used to read the session status attributes of the session management module and the anchor batch number. When the session state attribute is locked and there is no system abnormality, an enable control signal is sent to the back-end execution unit; When at least one of the following occurs: the session ends or an anomaly is detected, the enable control signal is cut off, the product removal robot is controlled to execute the diversion path, the conveyor control box is controlled to stop and at least one of the diversion actions is controlled, and the finished product label printer is locked.

Images