Mold injection molding full-process automation system based on layered cooperative control

Abstract

The invention discloses a mold injection molding full-process automation system based on layered cooperative control, which comprises a central control layer, a module scheduling layer and an equipment execution layer, and clock synchronization among the layers is realized through an industrial Ethernet. The central control layer comprises a dynamic scheduling server and a digital twinborn verification module, the module scheduling layer comprises a mold intelligent scheduling module, an injection molding process control module and a product quality closed loop module, and the equipment execution layer comprises a manipulator cooperation unit and a sprue material processing unit. By constructing a three-layer cooperative control architecture of a central control layer, a module scheduling layer and an equipment execution layer and combining a modular hardware carrier and a cross-layer data closed loop, common industrial problems such as cooperative level splitting, mold control marginalization and abnormal response delay are systematically solved; and finally, the comprehensive benefits that the die changing efficiency is remarkably improved, the product reject ratio is greatly reduced, and the energy consumption is obviously reduced are achieved.

Description

Technical Field

[0001] This invention relates to the field of injection molding automation systems, and in particular to a fully automated injection molding system based on hierarchical collaborative control. Background Technology

[0002] With the advancement of technology, in the field of injection molding automation systems, injection molding automation systems, as core equipment in the manufacturing industry, have achieved optimization of single-machine equipment control in recent years through distributed numerical control (DNC) and Internet of Things (IoT) technologies. For example, some systems significantly improve the stability of the molding process by collecting parameters such as injection molding machine pressure and temperature in real time and using adaptive PID algorithms; other solutions utilize RFID technology to track the mold's lifecycle, effectively reducing manual maintenance costs. Existing technologies have initially solved equipment-level control problems, laying an important foundation for production automation.

[0003] However, in flexible manufacturing scenarios involving multiple varieties and small batches, existing technologies still suffer from several systemic defects. First, there is a fragmented collaborative hierarchy at the system architecture level: the lack of a real-time closed-loop mechanism between the central scheduling layer and the execution units leads to independent operation of modules such as mold preheating, AGV scheduling, and injection molding processes. Mold changeover times remain long, and resource contention in multi-machine collaboration results in a high conflict rate. Second, mold control is marginalized: existing methods generally focus on optimizing injection molding machine parameters, failing to treat the mold as a core control object and neglecting the physical coupling between its temperature and the injection molding process, resulting in significant product dimensional fluctuations. Third, the system's response to abnormal operating conditions is significantly delayed: especially in handling faults such as mold blockage, manual intervention is still relied upon, leading to a low success rate in resolving moderate to severe faults. The overall delay from quality anomaly identification to process parameter adjustment is quite noticeable. Summary of the Invention

[0004] Therefore, it is necessary to provide a fully automated mold injection molding process system based on hierarchical collaborative control to address the technical problem of how to enable high-precision collaborative operation of molds, injection molding machines, auxiliary equipment, etc.

[0005] A fully automated mold injection molding process based on hierarchical collaborative control is disclosed. The system adopts a hierarchical collaborative control architecture, including a central control layer, a module scheduling layer, and an equipment execution layer. The layers are synchronized by clock through industrial Ethernet.

[0006] The central control layer is used for overall scheduling and virtual-real verification, including mold-order matching and dynamic production scheduling based on optimization algorithms, and early warning through real-time data comparison;

[0007] The module scheduling layer is used to schedule molds, control processes, and provide quality feedback.

[0008] The device execution layer is used to receive instructions to operate the robotic arm in collaborative work and gate cleaning, complete actions according to instructions, and provide real-time feedback of sensor data.

[0009] The system forms a closed-loop control through top-down command transmission and bottom-up data feedback.

[0010] In one embodiment, the central control layer includes a dynamic scheduling server and a digital twin verification module. The dynamic scheduling server includes a mold-order matching engine module and a conflict resolution module.

[0011] The mold-order matching engine module is used to generate mold allocation instructions based on order parameters and the mold database through an optimization algorithm;

[0012] The conflict resolution module is used to receive equipment load rate and order urgency data, and output production sequence adjustment instructions through the fuzzy logic controller.

[0013] The digital twin verification module is used to collect physical system data in real time through standard protocols and compare it with a preset simulation model. When the deviation meets a preset threshold, an early warning is triggered.

[0014] In one embodiment, the module scheduling layer includes: a mold intelligent scheduling module, an injection molding process control module, and a product quality closed-loop module.

[0015] The intelligent mold scheduling module is used to receive the mold allocation instruction, control the positioning and handling of the mold, and pre-start temperature control to pre-control the mold temperature.

[0016] The injection molding process control module is used to receive sensor data, dynamically correct the holding pressure curve and barrel PID parameters, and adjust the cooling water channel solenoid valves in different zones.

[0017] The product quality closed-loop module is used to trigger a clamping force compensation command to the injection molding machine after identifying flash and / or shrinkage marks, and to correct the mold temperature.

[0018] In one embodiment, the device execution layer includes: a robotic arm coordination unit and a sprue material processing unit;

[0019] The robotic arm collaboration unit is used to receive the mold opening signal from the injection molding process control module and perform trajectory pre-planning to pre-plan the running trajectory of the mold.

[0020] The gate material processing unit is used to form a physical linkage with the robotic arm coordination unit to synchronously remove gate residue.

[0021] In one embodiment, the dynamic scheduling server further includes a mold lifecycle management module; the mold lifecycle management module is used to obtain the number of times the mold ID is bound and its maintenance records.

[0022] In one embodiment, the dynamic scheduling server further includes a mold preheating strategy module, which is used to predict the next usage time of the mold and send a preheating start command to the mold intelligent scheduling module.

[0023] In one embodiment, the injection molding process control module further includes an anomaly handling coordination unit, which is used to detect the mold cavity pressure and perform unblocking operations based on the mold cavity pressure.

[0024] In one embodiment, the mold intelligent scheduling module is used to control the positioning and handling of molds, including: controlling the AGV to handle molds through laser positioning and RFID interaction.

[0025] In one embodiment, the mold intelligent scheduling module is used for pre-starting temperature control to pre-control the mold temperature, including: linking the embedded heater to pre-start temperature control to pre-control the mold temperature.

[0026] In one embodiment, the module scheduling layer further includes an energy consumption dynamic optimization module, which is used to control the flow rate of the cooling water pump.

[0027] The aforementioned automated injection molding system based on hierarchical collaborative control employs a three-layer collaborative control architecture consisting of a central control layer, a module scheduling layer, and an equipment execution layer. Combined with modular hardware and a cross-layer data closed-loop mechanism, it systematically solves core industry problems such as fragmented collaborative layers, marginalized mold control, and delayed anomaly response. The system achieves precise matching between molds and orders through the central control layer and effectively eliminates resource contention among multiple machines through real-time data comparison and dynamic calibration strategies. At the module scheduling layer, molds are intelligently scheduled through process and quality feedback, achieving physical coupling between process and state by treating the mold as the core control object. At the equipment execution layer, robotic arms coordinate operations and gate cleaning according to received instructions, completing actions as instructed and providing real-time sensor data feedback, significantly improving fault response and handling capabilities. This system significantly improves mold change efficiency and product quality, reduces defect rates and energy consumption, and enables high-precision collaborative operation of molds, injection molding machines, and auxiliary equipment. Attached Figure Description

[0028] Figure 1 This is a schematic diagram of the system framework structure of a fully automated mold injection molding system based on hierarchical collaborative control in one embodiment.

[0029] Figure 2 This is a schematic diagram illustrating the hierarchical relationship of the digital twin verification module in a three-layer collaborative control architecture in one embodiment.

[0030] Figure 3 This is a schematic diagram of the decision surface plot of the Gaussian membership function in one embodiment;

[0031] Figure 4 This is a schematic diagram of the hardware deployment framework in the conflict resolution module of one embodiment;

[0032] Figure 5 This is a schematic diagram of the system layered architecture of the module scheduling layer in one embodiment;

[0033] Figure 6 This is a schematic diagram of the collaborative operation sequence of the mold intelligent scheduling module in one embodiment;

[0034] Figure 7 This is a schematic diagram of the cooperative network of a multi-parameter closed-loop control unit in one embodiment;

[0035] Figure 8 This is a schematic diagram of the visual inspection process of the product quality closed-loop module in one embodiment. Detailed Implementation

[0036] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to provide a thorough understanding of the present invention. However, the present invention can be practiced in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below. In the description of the present invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," "counterclockwise," "axial," "radial," and "circumferential," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the present invention.

[0037] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0038] In this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components, unless otherwise explicitly limited. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0039] In this invention, unless otherwise explicitly specified and limited, "above" or "below" the second feature can mean that the first feature is in direct contact with the second feature, or that the first feature is in indirect contact with the second feature through an intermediate medium. Furthermore, "above," "over," and "on top" of the second feature can mean that the first feature is directly above or diagonally above the second feature, or simply that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature can mean that the first feature is directly below or diagonally below the second feature, or simply that the first feature is at a lower horizontal level than the second feature.

[0040] It should be noted that when an element is referred to as being "fixed to" or "set on" another element, it can be directly on the other element or there may be an intervening element. When an element is considered to be "connected to" another element, it can be directly connected to the other element or there may be an intervening element. The terms "vertical," "horizontal," "upper," "lower," "left," "right," and similar expressions used herein are for illustrative purposes only and do not represent the only possible implementation.

[0041] This invention provides a fully automated mold injection molding process based on hierarchical collaborative control. The fully automated mold injection molding process based on hierarchical collaborative control adopts a hierarchical collaborative control architecture, including: a central control layer, a module scheduling layer, and an equipment execution layer. The central control layer, the module scheduling layer, and the equipment execution layer are synchronized by clock via industrial Ethernet.

[0042] The central control layer is used for overall scheduling and virtual-real verification, including mold-order matching and dynamic production scheduling based on optimization algorithms, and early warning through real-time data comparison.

[0043] The module scheduling layer is used to schedule molds, control processes, and provide quality feedback.

[0044] The equipment execution layer is used to receive instructions to operate the robotic arm for collaborative work and gate cleaning, complete actions according to instructions, and provide real-time feedback of sensor data.

[0045] The system forms a closed-loop control through top-down command transmission and bottom-up data feedback.

[0046] The aforementioned automated injection molding system based on hierarchical collaborative control employs a three-layer collaborative control architecture consisting of a central control layer, a module scheduling layer, and an equipment execution layer. Combined with modular hardware and a cross-layer data closed-loop mechanism, it systematically solves core industry problems such as fragmented collaborative layers, marginalized mold control, and delayed anomaly response. The system achieves precise matching between molds and orders through the central control layer and effectively eliminates resource contention among multiple machines through real-time data comparison and dynamic calibration strategies. At the module scheduling layer, molds are intelligently scheduled through process and quality feedback, achieving physical coupling between process and state by treating the mold as the core control object. At the equipment execution layer, robotic arms coordinate operations and gate cleaning according to received instructions, completing actions as instructed and providing real-time sensor data feedback, significantly improving fault response and handling capabilities. This system significantly improves mold change efficiency and product quality, reduces defect rates and energy consumption, and enables high-precision collaborative operation of molds, injection molding machines, and auxiliary equipment.

[0047] In one embodiment, the central control layer includes a dynamic scheduling server, which comprises a mold-order matching engine module and a conflict resolution module. The mold-order matching engine module generates mold allocation instructions based on order parameters and a mold database using an optimized algorithm. Preferably, the mold-order matching engine module generates mold allocation instructions based on order parameters and a mold database using an improved Hungarian algorithm. The conflict resolution module receives equipment load rate and order urgency data and outputs production sequence adjustment instructions through a fuzzy logic controller.

[0048] In one embodiment, the central control layer includes a digital twin verification module. This module collects physical system data in real time via a standard protocol and compares it with a preset simulation model. An alert is triggered when the deviation meets a preset threshold. Preferably, the digital twin verification module collects physical system data in real time via the OPC UA protocol and compares it with a Simulink simulation model. An alert is triggered when the deviation is greater than 5%. It should be noted that the physical system data includes the mold cavity pressure value collected by a pressure sensor and the mold temperature value collected by an infrared thermometer. Triggering an alert when the deviation is greater than 5% means that when the actual collected data of the mold cavity pressure value and / or mold temperature value deviates from the simulation data of the Simulink simulation model by more than 5%, an alert is triggered to remind the user to re-test the fully automated injection molding system.

[0049] Furthermore, the term "mold" as used throughout this invention does not refer to a single mold, but is a general term encompassing all molds involved in the automated injection molding process within the hierarchical collaborative control-based fully automated mold injection molding system of this invention.

[0050] In one embodiment, the module scheduling layer includes: a mold intelligent scheduling module, an injection molding process control module, and a product quality closed-loop module.

[0051] The intelligent mold scheduling module receives mold allocation instructions, controls mold positioning and handling, and pre-activates temperature control to pre-regulate mold temperature. Preferably, the intelligent mold scheduling module receives the mold allocation instructions, controls the AGV to handle the mold via laser positioning and RFID interaction, and activates the embedded heater to pre-activate temperature control.

[0052] The injection molding process control module receives sensor data, dynamically adjusts the pressure value inside the mold and the PID parameters of the barrel, and regulates the cooling water channel solenoid valves in different zones. Preferably, the injection molding process control module receives data from the pressure sensor and infrared thermometer, dynamically corrects the holding pressure curve and the barrel PID parameters, and regulates the cooling water channel solenoid valves in different zones. It should be noted that the holding pressure curve is a mathematical curve formed by the combined effects of various factors such as the mold temperature, the barrel feed rate, and the flow rate of the cooling water channel controlled by the solenoid valve, while maintaining a stable pressure value in the mold cavity. The barrel PID parameters are the core parameters used to control the barrel's operating state, and closed-loop control is achieved through a combination of proportional (P), integral (I), and derivative (D). For specific meanings, please refer to existing technologies, which will not be elaborated here.

[0053] The product quality closed-loop module is used to trigger a clamping force compensation command to the injection molding machine after identifying flash and / or shrinkage marks, and to correct the mold temperature. Preferably, the product quality closed-loop module is used to identify flash and / or shrinkage marks through an online vision system, trigger a clamping force compensation command to the injection molding machine, and simultaneously use laser measurement of dimensional deviation data to reverse-correct the expansion coefficient of the mold temperature control system.

[0054] Furthermore, the product quality closed-loop module uses laser measurement of dimensional deviation data to reverse-correct the expansion coefficient of the mold temperature control system, including the following steps:

[0055] Step 1: Online Laser Dimension Measurement and Data Acquisition: After the injection molding process, the product enters the online inspection station via a conveyor belt. A high-precision laser measuring instrument is triggered to rapidly scan and measure the product's key dimensions, acquiring actual dimensional data for multiple feature points. The measurement data is uploaded in real-time to the central quality analysis system of the dynamic scheduling server via a high-speed industrial network.

[0056] Step 2: Measurement Data Preprocessing and Tolerance Judgment: After receiving the raw measurement data, the dynamic scheduling server first performs data filtering to eliminate measurement noise caused by environmental vibrations and other interference factors. Then, it automatically compares the processed actual size values ​​with the standard size tolerance range of the product design drawings to accurately calculate the size deviation value and the direction of the deviation (positive or negative deviation) for each feature point.

[0057] Step 3: Thermal Expansion Correlation Analysis: Based on the thermal expansion coefficient characteristics of the product material and mold steel, the system establishes a mathematical model between dimensional deviations and temperature changes. The distribution pattern of out-of-tolerance data is analyzed to identify deviation patterns with significant thermal expansion characteristics (such as uniform shrinkage or asymmetric deformation), thus eliminating accidental out-of-tolerance errors caused by non-temperature factors.

[0058] Step 4: Mold Temperature Field Reconstruction and Analysis: The system retrieves historical data from the mold temperature sensors during the production cycle, including the set temperature, actual temperature curves, and distribution of each zone heater. Combined with the mold's thermodynamic model, the system reconstructs the mold temperature field distribution during the molding process, identifying abnormal temperature areas or uneven distribution phenomena.

[0059] Step 5: Expansion Coefficient Compensation Calculation: Based on thermodynamic theory and historical data, a quantitative relationship model between dimensional deviation and mold temperature is established. Based on the currently measured dimensional deviation value, the required mold temperature adjustment amount and direction for compensating for this deviation are calculated in reverse. The nonlinearity of material shrinkage and the differences in thermal expansion in different areas of the mold are considered during the calculation.

[0060] Step 6: Adaptive Adjustment of Temperature Control Parameters: The system converts the calculated temperature compensation into specific adjustment parameters for the mold temperature control system. These parameters include: the set temperature adjustment value for a specific heating circuit, the temperature gradient adjustment scheme between each zone, and the optimized adjustment of the heating / cooling rate.

[0061] Step 7, Closed-Loop Verification and Iterative Optimization: In the next production cycle, the system uses the adjusted temperature parameters for production. After the product is formed, laser dimension measurement is performed again, and the new dimension data is compared with the expected improvement effect. Through data accumulation and analysis over multiple production cycles, the system continuously refines and optimizes the temperature compensation model, achieving continuous improvement.

[0062] Step 8: Knowledge Base Update and Experience Accumulation: Successful amendment examples and related parameters are saved to the process knowledge base of the dynamic scheduling server, forming temperature compensation experience data for specific molds, materials, and products. This data can provide initial parameter setting suggestions for the production of similar products in the future, enabling the system to learn and accumulate experience.

[0063] In this way, by establishing a direct correlation between the final product quality characteristics (dimensional accuracy) and process parameters (mold temperature) and achieving automated reverse correction, a true closed-loop quality control system is established. It can not only promptly correct deviations in current production but also improve the stability and intelligence of the entire production system through continuously accumulated correction experience.

[0064] It should be noted that the online vision system, clamping mechanism, laser measurement equipment, and mold temperature control system are pre-installed in the fully automated mold injection molding process based on hierarchical collaborative control, and are participants in this system just like the mold itself. This invention only aims to design a fully automated mold injection molding system based on hierarchical collaborative control. Due to space limitations, this invention only clarifies the hierarchical collaborative control mechanism and principle of the central control layer, module scheduling layer, and equipment execution layer. The automated mold injection molding system can be fully constructed within the existing technology framework, and this system solution can be directly called upon. The specific construction of the automated mold injection molding system can be found in existing technologies, and will not be elaborated upon here.

[0065] In one embodiment, the device execution layer includes a robotic arm coordination unit and a sprue handling unit. The robotic arm coordination unit receives the mold opening signal from the injection molding process control module and performs trajectory pre-planning to pre-plan the mold's running trajectory. The sprue handling unit is physically linked with the robotic arm coordination unit to synchronously remove sprue residue. Preferably, the sprue handling unit is physically linked with the robotic arm coordination unit to synchronously remove sprue residue using a negative pressure suction nozzle.

[0066] In one embodiment, the dynamic scheduling server further includes a mold lifecycle management module; this module is used to obtain the usage count and maintenance records associated with the mold ID. Preferably, the mold lifecycle management module obtains the usage count and maintenance records associated with the mold ID through an RFID reader. The RFID reader is highly stable and reliable.

[0067] In one embodiment, the dynamic scheduling server further includes a mold preheating strategy module. This module predicts the next usage time of the mold and sends a preheating start command to the mold intelligent scheduling module. Preferably, the mold preheating strategy module predicts the next usage time based on a historical molding cycle database and sends a preheating start command to the mold intelligent scheduling module. It is worth noting that the mold preheating strategy module generates a historical molding cycle database based on data such as mold usage frequency, number of uses, number of customer orders, and dynamic production scheduling history. This historical molding cycle database is a continuously learning, correcting, and enriching database. This embodiment involves collaborative mold lifecycle management, which uses mold ID binding to store mold data, RFID tags to store mold usage counts and maintenance records, and a central system that automatically triggers maintenance reminders and locks available machines. Under the mold preheating strategy, based on historical molding cycle data, the next usage time is predicted and preheating is started in advance, improving production efficiency.

[0068] Furthermore, since the calculation of preheating time and movement time requires information about the mold and the target injection molding machine after the mold intelligent scheduling module sends the preheating start command, and the equipment execution layer provides feedback on the position of the injection molding machine and the position of the mold, the mold-order matching engine module generates mold allocation instructions based on order parameters and the mold database using an improved Hungarian algorithm, including the following steps:

[0069] Step 1: Calculate the preheating time for each mold for each order: If the current temperature of the mold reaches the temperature required by the order (depending on the material), the preheating time is 0; otherwise, calculate the preheating time based on the temperature difference.

[0070] Step 2: Calculate the travel time from each mold to the injection molding machine specified for each order (based on AGV speed, distance, etc.).

[0071] Step 3: Construct an m*n cost matrix, where m is the number of orders and n is the number of molds.

[0072] Step 4: Solve using the Hungarian algorithm.

[0073] Furthermore, the mold-order matching engine module generates mold allocation instructions based on order parameters and the mold database using an improved Hungarian algorithm, including the following steps:

[0074] Step 1: Multi-dimensional Parameter Acquisition and Preprocessing: The system collects order parameters and mold status data in real time. Order parameters include material type, viscosity index, dimensional tolerance requirements, required mold temperature, priority weight, and delivery deadline. Mold status data covers material compatibility list, precision capability, current temperature, real-time geographical location, wear coefficient, maintenance history, and current energy consumption status. Simultaneously, the injection molding machine's operating status and geographical location information are acquired. All data undergoes normalization processing to eliminate dimensional differences, preparing for cost calculation.

[0075] Step 2: Dynamic Weighting Coefficient Calculation: The weight ratio of each cost item is dynamically adjusted based on the urgency and priority of the order. For orders with tight delivery deadlines, the weighting coefficients for material compatibility and precision matching are increased; for lower priority orders, the weighting ratio of energy consumption costs is appropriately increased. A time decay factor is introduced into the weighting calculation; the closer to the delivery deadline, the greater the impact of time urgency on the weighting.

[0076] Step 3: Multi-objective cost matrix construction: Calculate the comprehensive matching cost for each order-mold pair. First, determine material compatibility; if compatible, the cost is zero; otherwise, it is set to the maximum cost value. Then, calculate the precision matching degree; the ratio of precision difference to tolerance requirement is used as the precision cost. Next, calculate the temperature difference cost, considering the difference between the current mold temperature and the order's required temperature; differences exceeding the buffer temperature range are proportionally discounted. Simultaneously, calculate the mechanical wear cost, comprehensively evaluating the mold condition based on the wear coefficient and maintenance frequency. Finally, calculate the logistics cost, estimating the handling time by using the minimum distance between the mold and a compatible injection molding machine.

[0077] Step 4, Constraint Handling: Ensure that each order can only be assigned to one mold, and each mold can be assigned to at most one order. Check temperature feasibility, excluding matching pairs where the difference between the current temperature and the target temperature exceeds the heating capacity limit. Verify equipment compatibility, considering only molds that can be installed on available injection molding machines. Confirm mold availability status, excluding molds that require immediate maintenance or are under repair.

[0078] Step 5: Optimization and Allocation: The Hungarian algorithm is used to find the optimal allocation for the cost matrix. The algorithm simplifies the matrix through row and column reduction, and then uses the minimum line cover method to find independent zero elements. When no complete allocation can be found, the number of zero elements is increased through matrix transformations. Finally, the optimal allocation scheme is obtained, which minimizes the total matching cost.

[0079] Step 6: Preheating Plan Generation: Calculate the required preheating time for each mold based on the allocation results. Using a thermodynamic model, estimate the preheating time based on the current mold temperature, target temperature, heat capacity parameters, and heater power. Considering production line scheduling, if the preheating time exceeds the next available time window, initiate auxiliary heating measures or reassess the allocation plan.

[0080] Step 7, Logistics Instruction Generation: Generate AGV scheduling instructions based on the mold allocation results. Calculate the optimal transport path, considering the current mold position, the target injection molding machine position, and the status of the factory's logistics channels. Generate mold clamping instructions, including positioning coordinates, clamping posture, and transport speed parameters. Simultaneously, generate injection molding machine preparation instructions, including mold installation parameter settings and temperature pre-adjustment.

[0081] Step 8: Real-time Feedback and Adjustment: Monitor the execution of the allocation plan, and collect data on the actual preheating effect of the mold and the transportation progress. If the actual temperature does not meet expectations or transportation is delayed, trigger the reallocation mechanism in a timely manner. Dynamically update the cost matrix based on changes in production line status, and rerun the allocation algorithm when new orders are added or equipment malfunctions. Record the actual effect of each allocation for optimizing weight parameters and the cost model.

[0082] Step 9, Lifecycle Management: Update mold usage records, including usage count, cumulative working time, and most recent usage time. Predict mold maintenance needs based on usage data, and automatically exclude molds from the selection pool when maintenance thresholds are reached. Record energy consumption data to provide a foundation for energy efficiency optimization. Analyze matching quality data to continuously optimize cost calculation models and weight allocation strategies.

[0083] Thus, through steps 1 to 9 above, multi-objective optimization and real-time status perception are used to achieve high efficiency and economy in mold allocation, significantly improving mold utilization and energy efficiency compared to traditional methods.

[0084] Furthermore, the intelligent mold scheduling module is used to receive the mold allocation instruction, control the AGV to transport the mold through laser positioning and RFID interaction, and link the AGV intelligent transport and mold temperature control linkage method with the embedded heater pre-start temperature control, which includes the following steps:

[0085] Step 1: Task Instruction Issuance and Path Planning: The dynamic scheduling server issues transport tasks to the designated AGVs based on the allocation instructions generated by the mold-order matching engine. The task instructions include the unique identifier of the target mold, the destination injection molding machine number, and the priority. After receiving the instructions, the AGV control system plans the optimal path from its current location to the mold storage warehouse and then to the target injection molding machine based on a high-precision map of the factory area and real-time traffic conditions, while avoiding dynamic obstacles and congested areas.

[0086] Step 2, Mold Identification and Initial Positioning: The AGV travels to the designated area of ​​the mold storage warehouse. By reading the RFID tags placed on the ground, the AGV confirms that it has entered the precise operating area. Subsequently, the RFID reader on the AGV actively scans the RFID tags on the molds to obtain the mold's unique ID information and compares it with the task instructions to confirm that the target has been correctly grasped. This step completes the coarse positioning of the target, improving the AGV's navigation accuracy from ±10cm to ±2cm.

[0087] Step 3, Laser Scanning and Precise Positioning: After the target is confirmed by RFID, the AGV activates its onboard 2D laser scanner to perform a high-frequency scan of the mold and its surrounding environment. The point cloud data obtained from the scan is matched with the pre-stored 3D model of the mold using an iterative nearest-point algorithm to calculate the precise pose deviation of the mold relative to the AGV robotic arm interface (including X and Y coordinates and rotation angle θ).

[0088] Step 4, Pose Compensation and Precise Grasping: The AGV control system converts the calculated pose deviation value into motion compensation commands for each joint of the robotic arm. The robotic arm adjusts the grasping path of the end effector according to the compensation commands, completing the grasping and locking of the mold with millimeter-level precision. After successful grasping, the AGV sends a "Mold in place" signal back to the central system.

[0089] Step 5, Temperature Control Command Synchronous Trigger: The instant the robotic arm successfully grasps the mold, the AGV rewrites the mold tag using an RFID reader, marking its status as "transporting". Simultaneously, this action triggers a linkage command: the AGV sends the mold ID and destination injection molding machine ID to the central temperature control system of the dynamic scheduling server via a wireless network (such as 5G or Wi-Fi 6).

[0090] Step 6, Heating Strategy Matching and Activation: After receiving the instruction, the central temperature control system immediately queries the historical process database of the mold to obtain the optimal preheating temperature curve for the materials required for the current order. The system then issues a pre-start instruction to the embedded heater on the target injection molding machine, which includes key parameters such as the target temperature and heating rate.

[0091] Step 7, Status Synchronization During Handling: The AGV begins transporting the mold to the injection molding machine along the planned path. Simultaneously, the embedded heaters on the injection molding machine activate, heating up according to instructions. Real-time mold temperature data is transmitted back to the temperature control system via temperature sensors, enabling the central system to monitor the preheating progress.

[0092] Step 8, Arrival Confirmation and Final Positioning: The AGV transports the mold to the vicinity of the target injection molding machine. By reading the RFID marker in front of the injection molding machine, the AGV confirms that it has reached the correct workstation and reactivates the laser scanner to perform final position calibration with the fixed reference point on the injection molding machine, ensuring that the docking accuracy between the mold and the injection molding machine's mold closing mechanism meets the requirement of ±0.1mm.

[0093] Step 9: Mold Installation and Temperature Closed-Loop Switching: The AGV-controlled robotic arm precisely installs the mold onto the injection molding machine. A successful installation signal is sent back to the central system. At this point, the temperature control system seamlessly switches control of the mold temperature from the AGV-linked preheating system to the high-precision temperature control unit of the injection molding machine itself, forming a closed-loop control to ensure that the mold reaches and stabilizes at the required process temperature before mold closing.

[0094] Step 10, Task Closure and Feedback: After the AGV is released, it moves out of the work area and reports task completion to the dynamic scheduling server. Simultaneously, the injection molding machine confirms that the mold temperature has reached the set value and production conditions are met, then sends a "ready" signal to the central system. At this point, the entire collaborative operation from material handling to temperature control preheating is complete, and the system awaits the injection molding start command.

[0095] Thus, through steps 1 to 10 above, coarse positioning and status tracking are achieved through information flow (RFID), fine positioning and motion compensation are achieved through data flow (laser point cloud), and the pre-action of the physical execution unit (heater) is triggered through control flow (wireless command). Ultimately, the efficient collaboration between the logistics and temperature control systems in time and space is achieved, significantly reducing mold preparation time and energy consumption.

[0096] In one embodiment, the injection molding process control module further includes an anomaly handling coordination unit, which is used to detect the mold cavity pressure and perform unblocking operations based on the mold cavity pressure. Preferably, when the pressure sensor detects an abnormal mold cavity pressure, the anomaly handling coordination unit extends the holding pressure stage, triggers the mold micro-vibration unit, visually confirms the unblocking result, and if it fails, sends a mold removal command to the mold intelligent scheduling module. As mentioned above, the holding pressure stage is the time period during which the pressure value inside the mold under the holding pressure curve remains stable. Extending the holding pressure stage when the pressure sensor detects an abnormal mold cavity pressure allows time for subsequent visual confirmation of the unblocking result.

[0097] In one embodiment, the module scheduling layer also includes an energy consumption dynamic optimization module, which controls the flow rate of the cooling water pump. It is understood that in an automated mold injection molding system, the cooling water pump's role is merely to pump water for cooling the mold, and the water flow rate is controlled by various solenoid valves, which are typically deployed in a matrix configuration.

[0098] To further illustrate the operating mechanism and principle of the fully automated mold injection molding system based on hierarchical collaborative control in the above embodiments, specific embodiments are provided below. It should be reiterated that the system of this invention is innovatively implemented using a hierarchical collaborative control mechanism, based on the premise that a fully automated mold injection molding system can be completely constructed under current technology and industry conditions.

[0099] Please see Figure 1 In one specific embodiment, the fully automated mold injection molding process system based on hierarchical collaborative control includes: a central control layer, a module scheduling layer, and an equipment execution layer. The central control layer, the module scheduling layer, and the equipment execution layer are synchronized by clock via industrial Ethernet. The industrial Ethernet is based on the TSN protocol, i.e., a time-sensitive network protocol, and the clock synchronization deviation is <1μs.

[0100] The central control layer includes a dynamic scheduling server and a digital twin verification module. The dynamic scheduling server integrates a mold-order matching engine module and a conflict resolution module. The mold-order matching engine module generates mold allocation instructions based on order parameters such as material viscosity and dimensional tolerances, and mold database parameters such as usage frequency and wear coefficient, using an improved Hungarian algorithm. The conflict resolution module receives equipment load rate and order urgency data, and outputs production sequence adjustment instructions through a fuzzy logic controller.

[0101] The digital twin verification module collects physical system data in real time via the OPC UA protocol and compares it with the Simulink simulation model. When the deviation is greater than 5%, an early warning is triggered.

[0102] The module scheduling layer includes: a mold intelligent scheduling module, an injection molding process control module, and a product quality closed-loop module. The mold intelligent scheduling module receives the mold allocation instructions, controls the AGV to transport the mold via laser positioning and RFID interaction, and links the embedded heater to pre-start the temperature control. The injection molding process control module receives data from pressure sensors and infrared thermometers, dynamically corrects the holding pressure curve and barrel PID parameters, and adjusts the cooling water channel solenoid valves in different zones. The product quality closed-loop module identifies flash / shrinkage marks through an online vision system, triggers clamping force compensation instructions to the injection molding machine, and uses laser measurement of dimensional deviation data to reversely correct the expansion coefficient of the mold temperature control system.

[0103] The equipment execution layer includes a robotic arm coordination unit and a gate material processing unit. The robotic arm coordination unit is used to receive the mold opening signal from the injection molding process control module and execute trajectory pre-planning. The gate material processing unit is used to physically link with the robotic arm coordination unit and synchronously remove gate residue through a negative pressure suction nozzle.

[0104] Real-time data from each sensor in the device execution layer is fed back to the digital twin verification module. After verification by the digital twin verification module, the verification data is sent to the dynamic scheduling server. The dynamic scheduling server issues dynamic scheduling instructions to the module scheduling layer, and the module scheduling layer sends device control instructions to the device execution layer. Real-time data from the sensors in the device execution layer is fed back to the dynamic scheduling server to form a closed loop.

[0105] The aforementioned automated mold injection molding process system based on hierarchical collaborative control systematically solves the three major industry challenges in the background technology by constructing a three-layer collaborative control architecture of central control layer - module scheduling layer - equipment execution layer, combined with modular hardware carrier and cross-layer data closed loop:

[0106] First, the problem of fragmented collaboration levels has been significantly improved. The central control layer achieves precise matching between molds and orders through a dynamic scheduling server, and dynamically calibrates the scheduling strategy with the help of real-time optimization algorithms and digital twin verification technology, effectively eliminating resource conflicts in the multi-machine collaboration process, greatly shortening mold changeover time, and significantly reducing the conflict incidence rate.

[0107] Secondly, a breakthrough has been achieved in addressing the issue of marginalized mold control. This system treats the mold as the core control object, achieving real-time coupling between the injection molding process and the mold's state through a module scheduling layer. The product quality closed-loop module dynamically adjusts the mold temperature control system based on visual and dimensional inspection data, significantly narrowing product dimensional fluctuations and noticeably improving control precision.

[0108] Third, the problem of delayed response to anomalies has been effectively mitigated. The equipment execution layer has significantly improved its response speed to abnormal operating conditions through real-time signal recognition and pre-action mechanisms. Combined with collaborative processing mechanisms such as mold self-unblocking, the system's success rate and timeliness in handling moderate faults have been significantly improved.

[0109] In summary, this hierarchical collaborative control system significantly improves mold-changing efficiency, greatly reduces product defect rate and energy consumption, effectively breaks through the bottleneck of flexible manufacturing, conforms to the development direction of Industry 4.0 cloud-edge collaboration, and has reliable industrial feasibility.

[0110] It should be noted that the hardware component of the central control layer mainly includes a dynamic scheduling server and a digital twin verification module. The dynamic scheduling server, as a physical device, is primarily used to run matching engines and conflict resolution algorithms. The digital twin verification module can also be considered an independent functional layer, typically existing as an edge computing node in actual deployments. By integrating commonly used digital twin platforms with industrial IoT sensor networks, it enables real-time acquisition of various sensor data from the device execution layer and full-process virtual debugging. The specific location and connection relationships of this module in the three-layer collaborative control architecture can be found in [link to relevant documentation]. Figure 2In a real-world configuration, this module runs on a high-performance edge computing unit, such as the NVIDIA Jetson AGX Orin edge computing unit. It employs a high-precision clock synchronization protocol to ensure timing consistency across all levels and achieves smooth real-time simulation results.

[0111] The aforementioned three-layer collaborative control architecture-based automated mold injection molding system directly addresses three major bottlenecks currently facing the industry: first, the lack of cross-layer closed-loop coordination between central decision-making, equipment execution, and quality verification; second, insufficient real-time collaboration between mold dynamic scheduling and injection molding processes; and third, the lack of millisecond-level response unit-level micro-collaboration mechanisms, such as robot pre-start and zoned cooling functions. These bottlenecks directly contribute to the industry's common problems of high overall system energy consumption and persistently high product defect rates, as detailed in international industry reports. The layered collaborative control system proposed in this invention achieves high-precision collaborative operation of molds, injection molding machines, and auxiliary equipment, as shown in Table 1 below.

[0112]

[0113] Table 1

[0114] Table 1 above addresses how the technical solution of this invention directly tackles the three major bottlenecks that the industry urgently needs to overcome, from three aspects: technical problems, closed-loop path, and solution effect. It clarifies that this invention constructs a "three-layer collaborative control architecture" (central control layer - module scheduling layer - equipment execution layer), achieving high-precision collaboration between molds, injection molding machines, and auxiliary equipment through dynamic task allocation algorithms and real-time data closed-loop feedback.

[0115] Specifically, as the central control layer at the global optimization level, its functions include production order parsing, dynamic resource scheduling, and energy efficiency optimization. Its coordination mechanism involves constructing a mold-order matching engine (mold-order matching engine module) to automatically match available molds in the mold library based on order parameters (material, size) and generate mold preheating instructions. A conflict resolution algorithm (conflict resolution module) is constructed to detect in real-time when multiple injection molding machines compete for molds or logistics resources, and dynamically adjusts the production sequence (e.g., fuzzy logic determination of order priority).

[0116] Furthermore, the central control layer, serving as the global optimization layer, incorporates a cross-layer collaborative verification mechanism—the digital twin verification module, also known as the digital twin verification layer. This module collects physical system data (pressure / temperature / position) in real time; performs millisecond-level comparisons with the Simulink simulation model, triggering an alert when the deviation exceeds 5%; and dynamically updates simulation parameters (such as the mold wear coefficient) to ensure model self-evolution. This achieves cross-layer data fusion: the central control layer's decision algorithm receives data from the verification layer and dynamically adjusts scheduling strategies (such as automatically reducing weight when the mold usage limit is exceeded), while the execution layer's robot arm motion trajectory data is fed back to the scheduling layer to optimize AGV (Automated Guided Vehicle) path planning. This solves the problem of "decision-execution disconnect" in traditional layered architectures.

[0117] It should be noted that the mold-order matching engine module generates mold allocation instructions based on order parameters and a mold database using an optimized algorithm. These instructions are generated using a dynamic task allocation algorithm, which, based on order parameters and the mold database, achieves precise matching between molds and orders through an improved optimization matching algorithm. With the support of commonly used digital twin simulation platforms, the core functionality of the mold-order matching engine module can be implemented as follows: The algorithm first comprehensively considers factors such as material compatibility, dimensional accuracy matching, and energy consumption costs to construct a comprehensive evaluation matrix; then, an optimized allocation algorithm is used to solve this matrix, automatically selecting the most suitable mold resources and generating mold allocation instructions. Verified in a multiphysics simulation environment, this matching method performs excellently in numerous virtual order tests, significantly improving mold utilization while effectively reducing preheating energy consumption.

[0118] Figure 3 The diagram illustrates the structure of the decision surface graph of the Gaussian membership function. Based on this, it can be understood that the conflict resolution algorithm of the conflict resolution module is implemented through a fuzzy logic control strategy, the core of which lies in the reasonable definition of input and output variables. The controller takes order urgency, mold preparation time, and equipment load status as inputs, and uses common simulation design tools to construct an inference mechanism with the help of continuous membership functions. The final output is a priority adjustment coefficient to dynamically coordinate production resources.

[0119] During the fuzzification process, the system achieves a soft partitioning of the actual production state by defining membership functions for the input variables. Taking order urgency as an example, this method divides it into multiple typical states and uses continuous membership functions to describe them, resulting in a smooth transition between different urgency levels. Compared to simple piecewise functions, this type of function better reflects the gradual changes in priority during production, effectively avoiding the abrupt changes that may occur in traditional rule bases. This ensures the continuity and coordination of the algorithm output, making it more in line with actual scheduling needs.

[0120] To better understand the execution logic of the conflict resolution module of this invention and the origin of the production sequence adjustment instructions, in this embodiment, the conflict resolution algorithm is implemented based on a fuzzy logic controller. Its core design lies in the reasonable definition of input and output variables. The system uses order urgency, mold preparation time, and equipment load status as inputs, constructs an inference mechanism through continuous membership functions, and finally outputs priority adjustment coefficients to achieve dynamic coordination of production resources. To fully illustrate the design of this fuzzy logic controller, the following section combines... Figure 3 and Figure 4 This explanation will be elaborated from four aspects: the technical essence, the interpretation of the decision surface, the feasibility assurance of the technical solution, and the simulation of specific implementation methods.

[0121] First, the technical essence of the fuzzy logic controller: The design of the fuzzy logic controller includes the definition of input and output variables, the fuzzification process, and the design of the rule base. Input variables include order urgency, mold preparation time, and equipment load rate, reflecting the urgency of order delivery, the preheating time for mold installation, and the current capacity utilization of the equipment, respectively. The output variable is the priority adjustment coefficient, used to dynamically adjust the production sequence. During the fuzzification process, the system uses continuous membership functions to partition the input variables into fuzzy sets, which better reflects the gradual change characteristics of priority in actual production and avoids the abrupt changes in traditional control. The rule base is designed based on actual production scenarios and contains multiple fuzzy rules. For example, it can significantly increase the priority for urgent orders when the mold is ready, or appropriately decrease the priority under high load conditions to avoid the risk of equipment overload.

[0122] Second, the decision surface reflects the nonlinear mapping relationship between input and output variables. Specific points in the surface correspond to high-priority adjustments, strong negative adjustments, and system innovation points: that is, proactively increasing the priority of tasks with long preparation times under low load conditions to make full use of idle equipment resources.

[0123] Third, the feasibility of the technical solution is guaranteed. The system adopts a standardized industrial real-time communication protocol to ensure high-precision time synchronization between components and extremely short single inference time, fully meeting the real-time scheduling requirements of the production line. In terms of algorithms, the system is built on a mature fuzzy inference library. By defining input and output variables and their fuzzy sets, combined with a manually corrected rule base, it achieves the mapping from multi-variable inputs to priority adjustment quantities. The defuzzification process uses common methods to ensure reasonable and continuous output, ultimately achieving dynamic scheduling optimization.

[0124] Fourth, a specific implementation simulation design was conducted. By simulating a scenario of sudden surge in orders and high equipment load, the processing effects of the traditional solution and this solution were compared. This solution uses a fuzzy controller to calculate the output of a strong negative adjustment result and dynamically selects alternative scheduling strategies, such as splitting orders and allocating them to standby equipment. Combined with a mold preheating strategy to compensate for preparation time, it ultimately achieves on-time order delivery and effectively controls equipment conflict rate, demonstrating the applicability and effectiveness of the algorithm in actual production.

[0125] In summary, this conflict resolution algorithm transforms the production scheduling problem into a multivariate nonlinear optimization problem through a fuzzy logic controller. Its technical contribution lies in using mold preparation time and equipment load rate as collaborative variables to effectively resolve resource contention conflicts. At the same time, it provides complete parameter design, rule base examples, and system interface solutions, ultimately achieving a significant reduction in equipment conflict rate. This demonstrates the entire process of solving technical problems and achieving technical effects through technical means.

[0126] like Figure 1 and Figure 5 As shown, the module scheduling layer includes: a mold intelligent scheduling module, an injection molding process control module, and a product quality closed-loop module. The module scheduling layer is the hub of the three-layer architecture, transforming central decisions into executable instructions for the equipment, while simultaneously collecting real-time data and feeding it back to the upper layer. The mold intelligent scheduling module, as the central hub for physical resource coordination, has input / output interfaces that meet the data parameters shown in Table 2 below:

[0127] Input source Output instructions Collaboration Objects Central layer mold allocation instructions AGV path planning AGV navigation system Mold RFID information Embedded heater start / stop command Mold temperature control unit Injection molding machine status signals robotic arm grasping posture correction Six-axis robotic arm

[0128] Table 2

[0129] Based on the parameters and indicators in Table 2 above... Figure 6 The intelligent scheduling module for molds can clearly show the real-time interaction process between each unit through its collaborative operation sequence diagram.

[0130] Furthermore, the module scheduling layer adopts a distributed control approach. The intelligent mold scheduling module includes an AGV + robotic arm collaborative positioning unit and a mold temperature control linkage unit. The AGV + robotic arm collaborative positioning unit controls the AGV to transport the mold to the injection molding machine. Through laser positioning and interaction with the machine's RFID, the robotic arm accurately grasps and installs the mold, achieving extremely high positioning accuracy. The mold temperature control linkage unit controls the activation of the embedded heating unit before mold installation, with temperature data fed back to the injection molding machine's barrel temperature control system in real time for synchronous preheating. In one embodiment, the positioning accuracy of the laser positioning is achieved by the AGV's SICK OD5000 laser. Thus, through laser positioning and interaction with the machine's RFID, the robotic arm accurately grasps and installs the mold, achieving high precision and efficiency.

[0131] The injection molding process control module includes a multi-parameter closed-loop control unit and a mold cooling coordination unit. The multi-parameter closed-loop control unit monitors the mold cavity pressure using a pressure sensor and dynamically adjusts the holding pressure curve accordingly. It also detects the melt temperature using an infrared thermometer and corrects the barrel heating PID parameters in real time. The coordination network of the multi-parameter closed-loop control unit can be found in [reference needed]. Figure 7 .from Figure 7 As can be seen, the multi-parameter closed-loop control unit optimizes the process through real-time monitoring and dynamic adjustment. This unit continuously collects data from infrared temperature and mold cavity pressure sensors and adaptively adjusts the injection molding process parameters accordingly. For example, when the melt temperature exceeds the set tolerance, the system automatically adjusts the heating control parameters; if the mold cavity pressure does not reach the set threshold, the holding time is extended accordingly to improve product quality. Simultaneously, the system can also adjust the opening of the cooling water valves according to the wall thickness characteristics of different areas of the product, achieving precise temperature control, thereby effectively reducing defects and improving molding consistency.

[0132] The mold cooling coordination unit is used to adjust the flow rate of the mold cooling channels in zones according to the product wall thickness distribution, significantly shortening the cooling time. The zoned adjustment of the mold cooling channel flow rate is controlled by a solenoid valve matrix. In one embodiment, the zoned cooling optimization of the injection molding process control module satisfies the following: the cooling channel flow rate is controlled by a solenoid valve matrix; when the product wall thickness is at a certain typical size, the cooling time is controlled within a preset time range, and correspondingly, the flow rate adjustment response delay is within a preset time range.

[0133] It is understandable that the main task of the product quality closed-loop module is to eliminate the root causes of quality defects. For this purpose, please refer to... Figure 8 Let's understand the technical closed-loop path of the product quality closed-loop module. In terms of unit collaboration, the product quality closed-loop module mainly adopts a visual inspection-process parameter linkage approach. Based on the online vision system, it identifies flash and shrinkage marks and automatically triggers injection molding machine clamping force compensation or holding pressure time adjustment accordingly. At the same time, it corrects the expansion coefficient of the mold temperature control system based on dimensional laser measurement deviations.

[0134] To better understand dimensional deviation reverse compensation, and thus to achieve more precise control over product dimensional deviations, the system employs a temperature reverse compensation mechanism based on the principle of thermal expansion. This function calculates the required mold temperature adjustment in real time based on the detected dimensional deviation value, combined with the material's coefficient of thermal expansion and the product's baseline dimensions. This adjustment is then optimized using a compensation gain coefficient, and finally, the adjustment command is sent to the mold temperature control system. This compensation strategy effectively reduces dimensional fluctuations caused by thermal deformation, significantly improves product dimensional consistency, and reduces the defect rate.

[0135] In this invention, the establishment of the equipment execution layer should respond in real time to the central control layer and the module scheduling layer. The equipment execution layer includes a robot arm coordination unit and a sprue material handling unit. The robot arm coordination unit controls the robot arm's micro-coordination in part picking. The injection molding machine's mold opening signal triggers the robot arm's motion trajectory pre-start, significantly shortening the part picking cycle. At the same time, while the robot arm is picking up the part, the negative pressure suction nozzle synchronously picks up the sprue material, avoiding residue from affecting mold closing, thus forming a synergy between the sprue material handling unit and the robot arm coordination unit.

[0136] In one specific embodiment, the robotic arm's part-picking operation employs a unit-level micro-cooperative algorithm based on motion trajectory pre-start. This algorithm monitors the mold opening angle of the injection molding machine in real time. When the mold opening angle exceeds a certain preset threshold, the robot's trajectory planning process is triggered in advance. The trajectory planning uses a high-order polynomial interpolation algorithm, which can generate smooth, continuous motion paths with reasonable mechanical characteristics under strict time constraints, effectively avoiding problems such as sudden movements and impacts, and achieving high-precision pre-positioning of the robot. Thus, through integrated simulation environment verification, the above-mentioned pre-start mechanism significantly shortens the overall part-picking operation cycle, improves the system response speed and coordination efficiency, thereby effectively reducing mold change time and enhancing the flexibility of the production line.

[0137] Furthermore, to avoid safety risks caused by over-reliance on the injection molding machine's mold opening signal, in one embodiment, the action triggering mechanism of the robotic arm collaborative unit is as follows: the injection molding process control module performs fast Fourier transform analysis on the mold cavity pressure data to identify characteristic points in the holding pressure stage; after the characteristic points are identified, a pre-instruction is sent to the robotic arm collaborative unit within a very short time, causing the part-removal action to start hundreds of milliseconds before the physical mold opening signal. In this embodiment, a pressure curve characteristic triggering mechanism is adopted. To clarify this mechanism, the workflow of the robotic arm's injection molding machine timing optimization scheme is explained in detail in five stages below:

[0138] First, in the data acquisition stage, the injection molding machine collects mold cavity pressure data in real time using a high-frequency sampling frequency. The pressure sensor transmits the raw data stream to the injection process control module via an industrial fieldbus. Its typical pressure curve characteristics are: the pressure rises rapidly during the injection stage, then enters a peak holding state, and finally enters a pressure holding stage with a gradual pressure decrease.

[0139] Secondly, in the feature extraction stage, the injection molding process control module performs sliding window processing on the pressure data stream and uses bandpass filtering to capture the end-of-pressure-holding characteristic. The energy percentage of a specific frequency band is extracted using frequency domain analysis. When this energy percentage falls below a set threshold, the end of the pressure-holding stage is determined and marked. Extensive simulation testing has demonstrated that this timing identification has extremely high accuracy.

[0140] Third, in the decision-making and execution phase, pre-action command generation: After feature point recognition, the system sends pre-action commands to the robot arm with a fixed delay. These commands include information such as action type, target pose, speed limits, and safety boundaries. Safety space verification: The robot arm's collaborative unit calculates the kinematic envelope in real time and executes a collision prediction algorithm to ensure the motion path remains within a safe range.

[0141] Fourth, in the pre-action execution phase, the trajectory is executed in segments: the robotic arm moves from the starting point through the safe intermediate point to the final pick-up position. During the pre-action phase, the robotic arm moves to the safe intermediate point at a set speed, with the range of motion of the joints always limited within the safety envelope. The entire process involves multiple key time nodes, and strict time synchronization is maintained between each node.

[0142] Fifth, in the mold opening linkage stage, when the injection molding machine triggers physical mold opening, the robot arm has already started the part-picking action and reached the safety point. Subsequently, the robot arm completes the remaining path movement and finally completes the part-picking operation.

[0143] The above five stages detail the entire working process of the injection molding machine timing optimization scheme for the robotic arm. To this end, the injection molding process control module employs a triple safety protection mechanism to achieve full-process safety monitoring: first, a spatial pre-verification mechanism for calculating the safety envelope before motion commands; second, a real-time millimeter-wave monitoring mechanism; and third, an immediate stop mechanism when joint torque exceeds a threshold. By predicting the mold opening timing in advance through spectrum analysis, it overcomes the limitations of traditional mold opening signal triggering; it completes most of the safety path before physical mold opening, significantly shortening part removal time; it generates a safety space in real time using the injection molding machine's kinematic model, achieving zero collision risk; and it adopts a high-precision clock synchronization protocol to ensure minimal system clock deviation.

[0144] To automate the process and replace manual intervention, avoiding the time-consuming task of operators judging the degree of blockage and trying different treatments in traditional solutions, one embodiment of the injection molding process control module includes an anomaly handling coordination unit. This unit is used to: extend the holding pressure stage, trigger the mold micro-vibration unit, and visually confirm the unblocking result when the pressure sensor detects an abnormal mold cavity pressure. If the unblocking fails, a mold removal command is sent to the mold intelligent scheduling module. Specifically: First, if the pressure sensor detects an abnormal curve, such as a sudden drop in pressure during the holding pressure stage, the system will not immediately alarm and stop, but will automatically extend the holding pressure time—giving the mold a chance to "self-repair." Second, the mold micro-vibration unit will only be activated if extending the holding pressure is still ineffective. The key here is that the vibration parameters must match the current injection molding material. For example, some materials require high-frequency, low-amplitude vibration, while others require low-frequency, high-amplitude vibration. During vibration, mold temperature changes are monitored to avoid damage to the mold due to overheating caused by friction. Finally, the visual confirmation step uses a high-speed industrial camera to judge the unblocking effect by comparing the pixel changes in the gate area before and after vibration. If the difference in grayscale values ​​is below the threshold, the unblocking process is deemed to have failed.

[0145] The processing flow of the aforementioned anomaly handling collaborative unit can be summarized into four consecutive steps, forming a complete automated response mechanism. First, the pressure in the mold cavity is monitored in real time by a pressure sensor. Once a significant deviation of the pressure curve from a preset threshold is detected, the system immediately triggers an anomaly signal, thus avoiding the significant delays associated with traditional manual observation. Second, in the initial handling stage, the system automatically extends the pressure holding time, continuously applying pressure to attempt to clear any potential blockages, effectively preventing production losses due to direct shutdowns. Subsequently, in the active intervention stage, the system activates the mold micro-vibration function, using vibrations at a specific frequency to dislodge blockages, replacing the previously time-consuming manual handling method of disassembling the mold. Finally, the gate status is detected by a vision system, and based on the unblocking results, the system automatically determines whether to resume normal production or remove the mold and add it to the maintenance queue, significantly reducing the risk of errors in manual judgment.

[0146] It's worth noting that throughout the entire process described above, the system sends the mold removal command directly to the intelligent mold scheduling module, rather than controlling only a single device. This means the system can automatically allocate a new mold to the current device while adding faulty molds to the maintenance queue, truly achieving "fault handling without affecting production cycle time." By incorporating physical intervention into the control closed loop, the mold status directly participates in system decision-making, effectively solving the technical problem of marginalized mold control. Simultaneously, this solution significantly reduces downtime, avoids mold damage that may result from manual intervention, and significantly improves the removal of blockage residues.

[0147] To achieve a real-time correction mechanism, in one embodiment, the digital twin verification module establishes a full-process injection molding simulation model through a professional simulation platform, receives temperature and pressure sensor data from the equipment's execution layer in real time, and compares it with the simulation results. When a deviation exceeds a set threshold, the module sends a parameter calibration command to the dynamic scheduling server. Essentially, the digital twin module constructs a two-way closed loop between the physical and virtual worlds: the full-process model established through the simulation platform is equivalent to configuring a "digital mirror" for the injection molding system. Its key advantage lies in real-time performance—ordinary digital twins are only used for monitoring, while this invention's module, leveraging standard industrial communication protocols, achieves rapid comparison between physical sensor data and virtual model data at millisecond speeds. Once the deviation exceeds a set tolerance, the system immediately triggers a parameter calibration command, enabling the dynamic scheduling server to adjust its strategy promptly. This type of real-time correction mechanism is precisely what is generally lacking in existing distributed CNC systems.

[0148] Of particular note is the point of technological coupling: the simulation model incorporates multi-physics coupling algorithms, including thermodynamics, fluid mechanics, and kinematics, such as a mold expansion coefficient prediction model. Calibration commands directly impact scheduling decisions. For instance, when the actual mold temperature is significantly higher than the model's prediction, the system automatically lowers the mold's allocation priority and triggers a maintenance reminder—this "perception-analysis-decision" closed loop is innovative in the injection molding field. While digital twins are already used in existing technologies, they are mostly limited to visual monitoring; this solution upgrades them to the core hub of cross-layer calibration, with the digital twin module acting as a cross-layer data center to verify the feasibility of scheduling commands in real time. When the simulation predicts a mold changeover time exceeding the expected duration, the system immediately calibrates the scheduling algorithm parameters, significantly shortening the actual mold changeover time. Simultaneously, by collecting real-time temperature data from the mold's embedded sensors and comparing it with the simulated expansion coefficient model, if the deviation exceeds a set threshold, process parameter calibration is triggered, achieving strict control over product dimensional fluctuations.

[0149] To achieve energy conservation and emission reduction, the preferred approach involves a dynamic energy consumption optimization module deployed at the module scheduling layer. This module controls the flow rate of the cooling water pump, adjusting it according to the mold temperature distribution curve. Simultaneously, it controls the hydraulic system to switch to a low-pressure variable displacement mode during the pressure holding phase, significantly reducing energy consumption during this stage. It is understood that the injection molding industry currently commonly uses an overall frequency conversion method, while this invention pioneers a "zoned on-demand energy supply and real-time load matching" mechanism. Two key points need to be emphasized in this invention: First, the temperature distribution curve drives flow regulation, i.e., a physical coupling mechanism. This invention uses an infrared thermal imager to scan the mold temperature field, generating a thermal map, which forms the temperature distribution curve. This curve then drives the solenoid valve matrix to adjust the flow in zones, achieving zoned frequency conversion cooling and eliminating the problem of overcooling in thin-walled areas. Second, the fundamental difference between variable displacement control during the pressure holding phase and conventional pressure regulation. In this invention, a pressure sensor detects the pressure holding phase and then adjusts and switches the variable displacement hydraulic pump in the hydraulic system, achieving variable displacement hydraulic control and further avoiding high-pressure redundancy.

[0150] In the above embodiments, the infrared thermal imager is synchronized with the solenoid valve control bus in real time; the variable displacement hydraulic pump uses high-performance hydraulic components; and the real-time flow matrix calculation is completed by a high-performance edge computing unit.

[0151] In one embodiment, the digital twin verification module runs on a high-performance edge computing unit, achieving inter-layer clock synchronization through a high-precision clock synchronization protocol and supporting high-frame-rate real-time simulation. This edge computing unit possesses powerful computing capabilities, enabling real-time simulation of multi-physics fields, a feat difficult to achieve with traditional industrial control equipment. Simultaneously, its high energy efficiency meets edge deployment requirements and its wide-temperature operating capability adapts to industrial environments. The key to the high-precision clock synchronization protocol lies in its extremely high time synchronization accuracy, which directly determines the closed-loop response speed of the three-layer architecture. Existing injection molding systems commonly use clock protocols with large synchronization errors, leading to delays in digital twin verification. This invention achieves extremely high-precision time synchronization through next-generation network switching technology. The set frame rate is not arbitrarily chosen but based on the computational requirements of the real-time simulation software; when the frame rate is too low, the fluid simulation error will exceed the allowable range. This edge computing unit maintains smooth real-time simulation performance even when running complex mold thermal models, fully meeting verification requirements. Furthermore, high-frame-rate simulation can predict AGV trajectory conflicts in real time, enabling precise control of mold change time.

[0152] In terms of industrial feasibility assurance, this invention prioritizes hardware selection. A high-precision SICK OD5000 laser positioner is used, and high-performance industrial-grade edge computing units such as the NVIDIA Jetson AGXOrin are selected. Secondly, regarding communication protocols, inter-layer communication employs real-time industrial communication protocols such as OPC UA over TSN (Time-Sensitive Networking), while inter-unit synchronous communication utilizes high-precision clock protocols such as the IEEE 1588 precision clock protocol with a deviation of <1μs. Finally, regarding safety mechanisms, primarily referring to the robot's safety mechanism, this involves establishing a pre-action safety zone for the robot and calculating the safety space envelope in real-time using the injection molding machine's kinematic model. Thus, compared to traditional automated injection molding systems, this invention offers shorter mold changeover times, significantly reduced overall energy consumption, a substantial increase in product yield, and record-breaking mold utilization.

[0153] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0154] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A fully automated mold injection molding process system based on hierarchical collaborative control, characterized in that: include: The system consists of a central control layer, a module scheduling layer, and an equipment execution layer, with clock synchronization achieved between these layers via industrial Ethernet. The central control layer is used for overall scheduling and virtual-real verification, including mold-order matching and dynamic production scheduling based on optimization algorithms, and early warning through real-time data comparison; The module scheduling layer is used to schedule molds, control processes, and provide quality feedback. The device execution layer is used to receive instructions to operate the robotic arm in collaborative work and gate cleaning, complete actions according to instructions, and provide real-time feedback of sensor data. The system forms a closed-loop control through top-down command transmission and bottom-up data feedback.

2. The fully automated mold injection molding process system based on hierarchical collaborative control according to claim 1, characterized in that, The central control layer includes a dynamic scheduling server and a digital twin verification module. The dynamic scheduling server includes a mold-order matching engine module and a conflict resolution module. The mold-order matching engine module is used to generate mold allocation instructions based on order parameters and the mold database through an optimization algorithm; The conflict resolution module is used to receive equipment load rate and order urgency data, and output production sequence adjustment instructions through the fuzzy logic controller. The digital twin verification module is used to collect physical system data in real time through standard protocols and compare it with a preset simulation model. When the deviation meets a preset threshold, an early warning is triggered.

3. The fully automated mold injection molding process system based on hierarchical collaborative control according to claim 2, characterized in that, The module scheduling layer includes: a mold intelligent scheduling module, an injection molding process control module, and a product quality closed-loop module. The intelligent mold scheduling module is used to receive the mold allocation instruction, control the positioning and handling of the mold, and pre-start temperature control to pre-control the mold temperature. The injection molding process control module is used to receive sensor data, dynamically adjust the pressure value inside the mold and the PID parameters of the barrel, and adjust the cooling water channel solenoid valves in different zones. The product quality closed-loop module is used to trigger a clamping force compensation command to the injection molding machine after identifying flash and / or shrinkage marks, and to correct the mold temperature.

4. The fully automated mold injection molding process system based on hierarchical collaborative control according to claim 3, characterized in that, The equipment execution layer includes: a robotic arm coordination unit and a sprue material processing unit; The robotic arm collaboration unit is used to receive the mold opening signal from the injection molding process control module and perform trajectory pre-planning to pre-plan the running trajectory of the mold. The gate material processing unit is used to form a physical linkage with the robotic arm coordination unit to synchronously remove gate residue.

5. The fully automated mold injection molding process system based on hierarchical collaborative control according to claim 4, characterized in that, The dynamic scheduling server also includes a mold lifecycle management module; the mold lifecycle management module is used to obtain the number of times the mold ID is bound and the maintenance records.

6. The fully automated mold injection molding process system based on hierarchical collaborative control according to claim 4, characterized in that, The dynamic scheduling server also includes a mold preheating strategy module, which is used to predict the next usage time of the mold and send a preheating start command to the mold intelligent scheduling module.

7. The fully automated mold injection molding process system based on hierarchical collaborative control according to claim 4, characterized in that, The injection molding process control module also includes an anomaly handling coordination unit, which is used to detect the mold cavity pressure and perform unblocking operations based on the mold cavity pressure.

8. The fully automated mold injection molding process system based on hierarchical collaborative control according to claim 4, characterized in that, The intelligent mold scheduling module is used to control the positioning and handling of molds, including controlling the AGV to handle molds through laser positioning and RFID interaction.

9. The fully automated mold injection molding process system based on hierarchical collaborative control according to claim 8, characterized in that, The intelligent mold scheduling module is used for pre-starting temperature control to pre-control the mold temperature, including: linking the embedded heater to pre-start the temperature control to pre-control the mold temperature.

10. The fully automated mold injection molding process system based on hierarchical collaborative control according to any one of claims 1 to 9, characterized in that, The module scheduling layer also includes an energy consumption dynamic optimization module, which is used to control the flow rate of the cooling water pump.

Images