Closed-loop control method and system for injection mold testing

Abstract

The invention provides an injection molding test mold closed-loop control method and system, and belongs to the technical field of injection molding. According to the method, a linear correlation model is called to reversely deduce a target screw stroke and a target back pressure value required by the theoretical target product weight based on the theoretical target product weight and the theoretical melt density. According to the method, the target screw stroke, the target backpressure value, the injection speed section number and the speed value of each section are combined, intelligent analysis on a simulation filling curve is combined, speed segmentation conforming to flow characteristics is automatically carried out, and the speed value of each section is determined, so that a mold testing process card for guiding mold testing operation can be generated on the basis of the target screw stroke, the target backpressure value, the injection speed section number and the speed value of each section in combination with actual production conditions. Therefore, the generated mold testing process card is not a result of experience trial and error any more, but a simulation digital body subjected to scientific modeling and feedback verification, field operation can be directly guided, and the problems of unstable product weight, mold testing blindness and the like caused by melt density fluctuation in traditional injection molding are fundamentally solved.

Description

Technical Field

[0001] This application relates to the field of injection molding technology, and in particular to a closed-loop control method and system for injection molding trial molding. Background Technology

[0002] Currently, Computer-Aided Engineering (CAE), as a mature theoretical analysis tool in the mold design stage, outputs core parameters such as filling time and pressure curves. However, in actual production, the melt density inside the barrel dynamically changes with the amount of material stored and the back pressure, leading to product weight fluctuations (approximately ±3%). This makes it difficult to directly and accurately convert simulation results into process parameters that can guide actual mold trials. This gap between theory and practice results in a blind approach to the mold trial process, high trial-and-error costs, and low efficiency, making it impossible to establish a scientific mold trial system where parameters can be mapped, problems can be traced, and results can be quantified. Summary of the Invention

[0003] To address at least one of the aforementioned problems, this application proposes a closed-loop control method and system for injection molding trial molding, which aims to directly convert CAE analysis results into executable trial molding process parameters through a linear correlation model (characterizing the linear correlation between product weight, screw stroke, and back pressure value), thereby achieving "first-time trial molding qualification".

[0004] To achieve the above objectives, a first aspect of this application proposes a closed-loop control method for injection molding trials, comprising: Obtain the theoretical filling parameters output by computer-aided engineering simulation, wherein the theoretical filling parameters include theoretical melt density and simulation filling curve; Based on the theoretical melt density and mold cavity volume, calculate the theoretical target product weight; The linear correlation model, which has been pre-trained using historical data, is invoked to back-calculate and optimize the target screw stroke and target back pressure value based on the theoretical target product weight and the theoretical melt density. The linear correlation model is used to characterize the linear correlation between product weight, screw stroke, and back pressure value. The number of injection speed segments is adaptively divided according to the shape of the simulated filling curve, and the speed value of each segment is calculated based on the injection time setting parameter. Based on the target screw stroke, the target back pressure value, the number of injection speed segments, and the speed values ​​of each segment, and in conjunction with actual production conditions, a trial molding process card is generated. The trial molding process card is used to guide the trial molding operation.

[0005] The above technical solution has the following advantages or beneficial effects: In this embodiment, the theoretical filling parameters output by computer-aided engineering simulation are first obtained, including the theoretical melt density and the simulated filling curve. Then, based on the theoretical melt density and mold cavity volume, the theoretical target product weight can be calculated. Next, a linear correlation model (characterizing the mapping relationship between product weight, screw stroke, and back pressure value) is invoked to inversely deduce the target screw stroke and target back pressure value required for the theoretical target product weight based on the theoretical target product weight and theoretical melt density. Combined with intelligent analysis of the simulated filling curve, speed segmentation conforming to flow characteristics is automatically performed, and the speed value of each segment is determined. Therefore, based on the target screw stroke, target back pressure value, number of injection speed segments, and speed values ​​of each segment, combined with actual production conditions, a trial molding process card is generated to guide trial molding operations. This makes the generated trial molding process card no longer the result of trial and error, but a simulated digital identity verified by scientific modeling and feedback. It can directly guide on-site operation and fundamentally solve the problems of unstable product weight (±3%) and blind trial molding caused by melt density fluctuations in traditional injection molding. This significantly reduces the frequency of repeated trial molding, shortens the development cycle, and establishes a scientific trial molding system with mappable parameters and traceable problems.

[0006] In one embodiment of this application, the step of calling a pre-trained linear correlation model based on historical data to back-calculate and optimize the target screw stroke and target back pressure value according to the theoretical target product weight and the theoretical melt density includes: The linear correlation model, which has been pre-trained using historical data, is invoked to calculate a set of estimated data based on the theoretical target product weight and the constraints of the current working conditions. The estimated data includes screw stroke and back pressure value. Obtain the actual weight of the product obtained by performing the injection under the conditions of the estimated data; The calculated density of the system is derived from the actual weight of the product, and the calculated density of the system is compared with the theoretical melt density. If the error between the system's calculated density and the theoretical melt density is greater than a set threshold, the linear correlation model is invoked to recalculate based on the theoretical target product weight and the constraints of the current working conditions, and the estimated data is updated until the error between the system's calculated density and the theoretical melt density is less than or equal to the set threshold. The screw stroke and back pressure values ​​at which the error between the system's calculated density and the theoretical melt density is less than or equal to a set threshold are taken as the target screw stroke and target back pressure values.

[0007] The above technical solution has the following advantages or beneficial effects: In this embodiment, the model's one-time calculation results are not directly used as the final instruction. Instead, the actual product weight is required after injection, and the calculated density is rigorously compared with the CAE theoretical density. When the error exceeds a set threshold (e.g., 13%), the system automatically triggers recalculation and parameter updates based on the same target weight (i.e., the theoretical target product weight) and current operating constraints, forming a feedback loop of "execution-measurement-verification-re-optimization". This design enables the system to have self-diagnosis and real-time correction capabilities, proactively responding to uncertainties such as material batch variations, equipment status fluctuations, or environmental interference. Ultimately, only when the density error stably converges to within the threshold will the corresponding parameters (including screw stroke and back pressure value) be locked as the target value. This not only eliminates the drawbacks of rigid parameters and inability to adapt to changes in operating conditions in traditional injection molding in principle, but also tightly couples CAE theory with actual production through a data-driven closed loop, thereby significantly improving the success rate of the first trial molding and the reliability of the process, achieving a qualitative leap from "experience-based trial and error" to "scientific closed-loop control".

[0008] In one embodiment of this application, the invocation of a pre-trained linear correlation model based on historical data calculates a set of estimated data according to the theoretical target product weight and the constraints of the current working conditions, including: Substitute the theoretical target product weight into the linear correlation model and establish the solution equation; Under the constraints of the current working condition, an optimization algorithm is used to select an optimal solution from all combinations of feasible solutions according to a preset optimization objective, so as to obtain a set of estimated data.

[0009] The above technical solution has the following advantages or beneficial effects: In this embodiment of the application, by substituting the theoretical target product weight into the linear correlation model and using an optimization algorithm to select the optimal parameter combination from an infinite number of feasible solutions under the constraints of the current operating conditions of the equipment, the injection molding trial can be completely transformed from an experience-dependent "open-loop trial and error" to a data-driven "closed-loop adaptive control".

[0010] In one embodiment of this application, the step of calculating the system's calculated density by back-calculating the actual weight of the product includes... Based on the actual weight of the product, the corresponding screw stroke, and the screw cross-sectional area, the calculated density of the system is derived.

[0011] The above technical solution has the following advantages or beneficial effects: In this embodiment, the density is calculated by back-calculating the actual product weight, the corresponding screw stroke, and the fixed screw cross-sectional area. This represents a fundamental shift from relying on uncontrollable theoretical density to closed-loop verification based on measurable physical quantities. Specifically, in traditional injection molding, density is considered an input parameter, and its fluctuations directly lead to uncontrolled product weight (±3%). This solution takes the opposite approach, utilizing the precisely measurable product weight and screw stroke—two stable and controllable mechanical quantities—to back-calculate density through geometric relationships. This makes "density" the sole objective benchmark for verifying the effectiveness of process parameters. When the error between the back-calculated density and the CAE theoretical density is ≤ a set threshold (e.g., 13%), it confirms that the combination of the current screw stroke and back pressure value can accurately reproduce the simulation expectation. Thus, without relying on manual experience or multiple trial moldings, the process card can be directly locked, achieving "one-time trial molding qualification." This mechanism completely bridges the data gap between simulation and production, upgrading trial molding from "experience-based trial and error" to "data-driven scientific verification," representing a key technological breakthrough for achieving precise and automated production in the fields of intelligent manufacturing and high-end equipment.

[0012] In one embodiment of this application, the adaptive division of the injection speed segments based on the morphology of the simulated fill curve includes: If the simulated filling curve is a single-peak curve, the injection speed segment will be adaptively divided into 3 segments; If the simulated filling curve is a bimodal or multimodal curve, the injection speed segment will be adaptively divided into 4 segments.

[0013] The above technical solution has the following advantages or beneficial effects: In this embodiment, when the CAE simulation filling curve shows a single-peak filling curve, it indicates that the melt flow is stable and the resistance is uniform. Only three speed control stages (such as low-speed gate filling, medium-speed filling of the main body, and high-speed holding pressure) are needed to efficiently complete the filling, avoiding parameter redundancy and debugging complexity caused by excessive segmentation. However, when the simulation filling curve shows a double-peak or multi-peak shape, it means that the melt front is experiencing backflow, merging, or stagnation in the cavity. This can easily lead to fatal defects such as air pockets, weld lines, or localized insufficient filling in the merging area. In this case, four speed control stages are necessary to precisely accelerate in critical areas (such as before merging) to eliminate trapped air, or to decelerate after merging to avoid overshoot and flash. This mechanism abandons the traditional fixed segmentation mode of "one-size-fits-all" and enables the system to intelligently select the most suitable control strategy based on the flow characteristics predicted by simulation. This ensures the molding reliability of complex structures and simplifies the process parameters of simple structures. It fundamentally transforms trial molding from "experience-based trial and error" to "morphology-driven scientific execution," making it an indispensable intelligent execution layer technology for achieving "first-time trial molding success."

[0014] In one embodiment of this application, after generating a trial molding process card based on the target screw stroke, the target back pressure value, the number of injection speed segments, and the speed values ​​of each segment, combined with actual production conditions, the method further includes: When a change of injection molding machine is detected, the current injection molding machine model is automatically identified, and the safe operating boundary of the current injection molding machine is retrieved from the preset parameter table; The process parameters in the generated trial molding process card are automatically checked against the safe operating boundary. When the process parameters exceed the corresponding safe operating boundary, they are automatically reduced and adjusted so that the process parameters do not exceed the corresponding safe operating boundary.

[0015] The above technical solution has the following advantages or beneficial effects: In this embodiment, after the trial molding process card is generated, the system automatically identifies the injection molding machine model and retrieves its safe operating boundaries (such as maximum injection pressure, maximum injection speed, rated back pressure range, etc.), and performs real-time verification and intelligent adjustment of the process parameters. This fundamentally achieves a qualitative change from "parameters usable" to "parameters safely usable," eliminating the risk of parameters exceeding limits due to differences in equipment capabilities. If the process card parameters exceed the physical limits of the new machine, the system does not require manual intervention and immediately compresses the parameters to a safe range without damage according to preset rules (such as prioritizing speed reduction and then adjusting pressure). This avoids equipment damage, mold deformation, or safety accidents caused by overpressure or overspeed, while preserving the process intent (such as filling shape and pressure holding effect) to the maximum extent, ensuring a continuous and stable trial molding process. This mechanism transforms the high-risk, high-cost process of "re-adjusting parameters by changing machines," which originally relied on engineers' experience, into a fully automated and standardized system behavior. This allows the same set of scientific trial molding process cards to be seamlessly transferred to injection molding machines of different brands, eras, and capacities, significantly reducing downtime during line changes and eliminating human error. It is a key safety cornerstone for building a flexible intelligent manufacturing system that allows for "one-time design and multi-machine compatibility."

[0016] In one embodiment of this application, after generating a trial molding process card based on the target screw stroke, the target back pressure value, the number of injection speed segments, and the speed values ​​of each segment, combined with actual production conditions, the method further includes: When the melt material is changed, the theoretical melt density corresponding to the changed material is loaded from the material database; The CAE simulation is retried based on the material change to generate the simulation fill curve corresponding to the material change. Returning to the step of calculating the theoretical target product weight based on the theoretical melt density and mold cavity volume, a trial molding process card corresponding to the material replacement is generated based on the theoretical melt density corresponding to the material replacement and the simulation filling curve corresponding to the material replacement.

[0017] The above technical solution has the following advantages or beneficial effects: In this embodiment, when the melt material is changed, the system automatically obtains the new melt density from the material database and re-simulates and generates a simulated filling curve reflecting its flow characteristics. Then, based on the updated theoretical target weight, it recalculates the screw stroke and back pressure combination to ensure that the injection speed segmentation strategy is fully adapted to the flow behavior of the new material. This process does not require manual intervention in CAE simulation or parameter adjustment, enabling the same mold to automatically generate a process card that strictly matches the material's physical properties (such as viscosity and flowability) when switching to different materials. This fundamentally avoids defects such as insufficient filling, weld lines, or warping caused by material differences, transforming the industry pain point of "mold adjustment required for material change" into a fully automated, zero-error intelligent process. This significantly improves production flexibility and product consistency, and is a key technological support for building a "one machine, multiple materials, one-click production change" intelligent manufacturing system.

[0018] To achieve the above objectives, a second aspect of this application provides a closed-loop control system for injection molding trials, comprising: The material quantity compensation algorithm module is used to perform the following steps: Obtain the theoretical filling parameters output by computer-aided engineering simulation, wherein the theoretical filling parameters include theoretical melt density and simulation filling curve; Based on the theoretical melt density and mold cavity volume, calculate the theoretical target product weight; The linear correlation model, which has been pre-trained using historical data, is invoked to back-calculate and optimize the target screw stroke and target back pressure value based on the theoretical target product weight and the theoretical melt density. The linear correlation model is used to characterize the linear correlation between product weight, screw stroke, and back pressure value. The screw curve optimization module is used to adaptively divide the number of injection speed segments according to the shape of the simulated filling curve, and calculate the speed value of each segment based on the injection setting time parameter. The process card generation module is used to generate a trial molding process card based on the target screw stroke, the target back pressure value, the number of injection speed segments, and the speed values ​​of each segment, combined with actual production conditions. The trial molding process card is used to guide the trial molding operation.

[0019] The above technical solution has the following advantages or beneficial effects: In this embodiment, the injection molding trial mold closed-loop control system constructs a fully intelligent mapping channel from CAE simulation to actual production through the coordinated operation of three major modules: a material quantity compensation algorithm, a screw curve optimization module, and a process card generation module. This enables a paradigm shift in injection molding trial molds from experience-driven to data-driven approaches. Specifically, the material quantity compensation algorithm module accurately back-calculates the screw stroke and back pressure value based on a linear correlation model. Combined with the screw curve optimization module's intelligent segmentation and speed calculation of the simulated filling curve, it ensures a complete match between the injection strategy and melt flow characteristics. Finally, the process card generation module outputs directly executable trial mold instructions. This mechanism not only eliminates the parameter disconnection problem caused by melt density fluctuations in traditional injection molding but also achieves dynamic optimization and automated generation of process parameters through modular design. This significantly reduces trial mold costs and cycles, improves molding quality stability, and provides key technical support for establishing a scientific and traceable intelligent manufacturing system.

[0020] In one embodiment of this application, the system further includes: The data acquisition module is used to collect the screw stroke, back pressure value, and actual weight of the product from the injection molding machine.

[0021] The above technical solution has the following advantages or beneficial effects: In this embodiment, the data acquisition module acquires key production data such as the screw stroke, back pressure value, and actual product weight of the injection molding machine in real time and with high accuracy. This provides a real and reliable physical feedback basis for the entire closed-loop control system, enabling data-driven closed-loop verification and dynamic optimization from "theoretical simulation" to "actual production." In traditional injection molding, parameter adjustments rely on manual experience and lack real-time data support, leading to blind and inefficient trial molding processes. This data acquisition module, however, continuously collects screw position, back pressure status, and product weight through high-precision sensors, allowing the system to monitor process execution in real time. This provides direct evidence for subsequent density back-calculation, error verification, and parameter iterative optimization. This mechanism not only ensures dynamic matching between process parameters and actual production conditions but also eliminates human error through data-driven methods, significantly improving trial molding accuracy and efficiency, and laying a solid data foundation for achieving the goal of "passing the trial molding on the first attempt."

[0022] In one embodiment of this application, the system further includes: The human-computer interaction terminal is used to display the process parameters of the trial molding process card and to show the segmented adjustment interface of the number of injection speed segments.

[0023] The above technical solution has the following advantages or beneficial effects: In this embodiment, the human-machine interface terminal intuitively displays the complete parameters of the trial molding process card (such as screw stroke, back pressure value, injection speed, etc.) and the segmented injection speed adjustment interface. This enables visualization and real-time interactive optimization of process parameters, breaking the limitations of the "black box operation" in traditional injection molding trials and significantly improving the accuracy and efficiency of process adjustments. Operators can clearly view the process card parameters through the terminal and directly fine-tune the injection speed on the segmented adjustment interface according to actual production needs (such as filling status, defect feedback), quickly optimizing the injection strategy without relying on experience guessing or repeated trials. Simultaneously, this interface supports real-time data linkage with the closed-loop control system, ensuring that adjusted parameters are immediately fed back to the production process, avoiding molding defects caused by parameter transmission errors. This design not only lowers the operational threshold, enabling non-professionals to efficiently complete process optimization, but also further improves the success rate and stability of trial molding through human-machine collaboration, providing convenient and reliable interactive support for "one-time trial success."

[0024] It should be understood that the above general description and the following detailed description are merely exemplary and do not limit this application. Attached Figure Description

[0025] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the specification, serve to explain the principles of this application.

[0026] Figure 1 This is a first flowchart of a closed-loop control method for injection molding trial molds provided in an embodiment of this application.

[0027] Figure 2 This is a flowchart illustrating the steps of calling a pre-trained linear correlation model based on historical data to back-calculate and optimize the target screw stroke and target back pressure value according to the theoretical target product weight and theoretical melt density, as provided in one embodiment of this application.

[0028] Figure 3 This application provides a flowchart of the steps for calling a pre-trained linear correlation model based on historical data to calculate a set of estimated data according to the theoretical target product weight and the constraints of the current working conditions.

[0029] Figure 4 This is a flowchart illustrating the steps of adaptively dividing the number of injection speed segments based on the morphology of the simulated filling curve, according to an embodiment of this application.

[0030] Figure 5 This is a flowchart of the steps to be executed after generating a trial molding process card based on the target screw stroke, target back pressure value, number of injection speed segments and speed values ​​of each segment, combined with actual production conditions, according to an embodiment of this application.

[0031] Figure 6 This is a second flowchart of a closed-loop control method for injection molding trial molds provided in an embodiment of this application.

[0032] Figure 7 This is a structural block diagram of an injection molding trial mold closed-loop control system provided in one embodiment of this application.

[0033] Figure label: 700. Plastic trial mold closed-loop control system; 710. Material quantity compensation algorithm module; 720. Screw curve optimization module; 730. Process card generation module; 740. Data acquisition module; 750. Human-machine interaction terminal. Detailed Implementation

[0034] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0035] It should be noted that although functional modules are divided in the device schematic diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first," "second," etc., in the specification, claims, and the aforementioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0036] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0037] In traditional injection molding trials, while computer-aided engineering (CAE) simulations can output theoretical filling parameters (such as theoretical melt density and simulated filling curves), the actual melt density in production is affected by dynamic changes in operating conditions such as material storage and back pressure, leading to significant fluctuations in product weight (approximately ±3%). This makes it difficult to directly translate simulation results into accurate trial molding process parameters. Engineers must rely on experience to repeatedly adjust screw stroke, back pressure, and injection speed, resulting in low trial molding efficiency, high costs, and potential defects such as incomplete filling and weld lines due to parameter discrepancies. It is impossible to establish a scientific trial molding system where parameters can be mapped and problems can be traced. The industry urgently needs a closed-loop control method that can seamlessly integrate CAE simulation parameters with actual production to bridge the gap between theory and practice and improve trial molding accuracy and efficiency.

[0038] Based on this, this application provides a closed-loop control method for injection molding trial molding. By using a linear correlation model (characterizing the linear correlation between product weight, screw stroke and back pressure value), the CAE analysis results are directly converted into executable trial molding process parameters, achieving "one-time trial molding qualification".

[0039] Reference Figure 1 , Figure 1 This is a first flowchart of an injection molding trial mold closed-loop control method provided in an embodiment of this application, including but not limited to steps S110 to S150.

[0040] Step S110: Obtain the theoretical filling parameters output by the computer-aided engineering simulation. The theoretical filling parameters include the theoretical melt density and the simulation filling curve.

[0041] In this step, it is necessary to first obtain the theoretical filling parameters output by computer-aided engineering simulation. These theoretical filling parameters are core data output by CAE simulation software (such as Moldflow and Moldex3D), used to describe the flow behavior and physical properties of the melt in the mold cavity. Theoretical filling parameters include theoretical melt density and simulated filling curves. The theoretical melt density (ρ_theory) represents the density value of the melt under ideal conditions and is a key input for calculating the weight of the target product. CAE, through material property databases (such as UDB files) and thermodynamic models, combined with boundary conditions such as mold structure and temperature field, can predict the density distribution of the melt during the filling process. The simulated filling curve reflects the trajectory of the melt front's position within the cavity over time, including dynamic information such as velocity, pressure, and temperature. This curve reveals the temporal characteristics of melt filling (such as single-peak, double-peak, or multi-peak morphology) and is the direct basis for subsequent injection speed segmentation strategies.

[0042] Step S120: Calculate the theoretical target product weight based on the theoretical melt density and mold cavity volume.

[0043] In this step, the theoretical target product weight (W_target = ρ_theory × mold cavity volume) is calculated using the theoretical melt density (ρ_theory) and mold cavity volume output by CAE simulation. The simulation results are then converted into the target weight for actual production, ensuring the scientific validity and reliability of the process parameter calculations.

[0044] Step S130: Call the pre-trained linear correlation model based on historical data, and back-calculate and optimize the target screw stroke and target back pressure value based on the theoretical target product weight and theoretical melt density. The linear correlation model is used to characterize the linear correlation between product weight, screw stroke and back pressure value.

[0045] In this step, a linear correlation model trained on historical production data (used to characterize the linear relationship between product weight, screw stroke, and back pressure) is invoked. Using the theoretical target product weight calculated by CAE as the sole input, and within the physical constraints of the equipment (the upper and lower limits of screw stroke and back pressure), all feasible parameter combinations that satisfy the weight target are solved in reverse. Based on preset optimization objectives (such as minimum energy consumption and minimum equipment wear), the optimal set of initial parameters is selected. Subsequently, the first injection is performed, and the actual product weight is collected. The density is calculated using a formula and compared with the theoretical density obtained from CAE until the density error converges to within a threshold. Finally, the screw stroke and back pressure values ​​that meet the verification conditions are locked as the "target parameters." This process achieves self-verification, self-correction, and self-convergence from theoretical targets to executable process parameters, ensuring the physical reliability and engineering feasibility of the parameters. It is the core decision engine for the closed-loop system to achieve "first-time trial success."

[0046] Specifically, refer to Figure 2 , Figure 2 This application provides a flowchart of the steps for calling a pre-trained linear correlation model based on historical data to back-calculate and optimize the target screw stroke and target back pressure value based on the theoretical target product weight and theoretical melt density, including but not limited to steps S210 to S250.

[0047] Step S210: Call the linear correlation model that has been trained in advance using historical data, and calculate a set of estimated data based on the theoretical target product weight and the constraints of the current working conditions. The estimated data includes screw stroke and back pressure value. Step S220: Obtain the actual weight of the product obtained by performing injection under the calculated data; Step S230: Calculate the system calculated density based on the actual weight of the product, and compare the system calculated density with the theoretical melt density; Step S240: If the error between the system's calculated density and the theoretical melt density is greater than a set threshold, the linear correlation model is called to recalculate based on the theoretical target product weight and the current working conditions, and the calculated data is updated until the error between the system's calculated density and the theoretical melt density is less than or equal to the set threshold. Step S250: The screw stroke and back pressure values ​​at which the error between the system-calculated density and the theoretical melt density is less than or equal to a set threshold are taken as the target screw stroke and target back pressure values.

[0048] This application embodiment constructs a closed-loop self-optimizing system driven by theoretical objectives and fed back by measured data. The system first calls a linear correlation model trained and validated using historical production data. This linear correlation model characterizes the linear relationship between product weight, screw stroke, and back pressure. This linear correlation module can be represented by the following Equation 1: (Equation 1); In Equation 1, Indicates product weight. Indicates the screw travel. This is the back pressure value. This represents the weighting coefficient corresponding to the screw stroke. This represents the weighting coefficient corresponding to the back pressure value. , , These are the model parameters obtained by training with historical production data.

[0049] This application uses the theoretical target product weight Wtarget output from CAE simulation as the sole input. Combined with the physical constraints of the current injection molding machine (such as the minimum / maximum allowable values ​​of screw stroke and the upper and lower limits of back pressure), it uses mathematical inverse kinematics to select an optimal initial calculation parameter (Lscrew, Pback) from an infinite number of feasible parameter combinations based on preset optimization objectives (such as prioritizing reducing back pressure to save energy or prioritizing increasing stroke to ensure filling). Subsequently, the system performs a complete injection under this parameter combination, with a high-precision weighing sensor collecting the actual weight Wactual of the molded product in real time. Then, using the formula ρcalc=Wactual / Lscrew×A (where A is the screw cross-sectional area, determined by the inherent parameters of the equipment), the system-calculated density ρcalc corresponding to this injection is derived and precisely compared with the theoretical melt density ρtheory provided by CAE. If the relative error between the two exceeds a preset threshold (13%), the system determines that the current parameter combination fails to effectively reproduce the theoretical physical state. It then triggers an iterative mechanism, using the same Wtarget and equipment constraints as input, to re-solve the linear model and generate a new set of parameters. This process is automatically and seamlessly repeated within the control system until the error between ρcalc and ρtheory calculated after a certain injection stabilizes within the preset threshold (13%). Finally, the system officially locks the set of screw stroke and back pressure values ​​that meet this verification condition as the "target screw stroke" and "target back pressure value," and writes them as unique and reliable process parameters into the trial molding process card. The entire process does not rely on manual intervention. Its essence is to use the measured results from the physical world to dynamically calibrate and verify the output of the theoretical model, transforming "theoretical prediction" into a "deterministically confirmed solution." This ensures the accuracy, robustness, and traceability of the parameter conversion from CAE simulation to actual production, making it the core decision engine for achieving the industrial-grade goal of "one-time molding, multiple calibrations, and one-time qualification."

[0050] Reference Figure 3 , Figure 3 This application provides a flowchart of the steps for calling a pre-trained linear correlation model based on historical data to calculate a set of estimated data according to the theoretical target product weight and the constraints of the current working conditions, including but not limited to steps S310 to S320.

[0051] Step S310: Substitute the theoretical target product weight into the linear correlation model and establish the solution equation; Step S320: Under the constraints of the current working condition, an optimization algorithm is used to select an optimal solution from all combinations of feasible solutions according to the preset optimization objective, so as to obtain a set of estimated data.

[0052] In this embodiment, a precise transformation from theoretical targets to executable process parameters is achieved through a combination of mathematical modeling and engineering optimization. Specifically, the system first substitutes the theoretical target product weight obtained from CAE simulation into a pre-trained linear correlation model to establish a solution equation with this weight as the right-hand side of the equation. Subsequently, under the constraints of the current operating conditions, such as the physical operating boundaries of the injection molding machine (e.g., minimum / maximum screw stroke, upper and lower back pressure limits), an optimization algorithm (e.g., linear programming) is used to select a unique optimal solution from infinitely many sets of parameter combinations that satisfy the equation, based on the preset engineering optimization objectives (e.g., minimizing back pressure to reduce energy consumption, or maximizing stroke to improve filling stability). The final screw stroke and back pressure values ​​are then output as calculated data. This process does not rely on manual trial and error, but rather ensures that the generated parameters meet both weight accuracy requirements and equipment executability through the collaborative calculation of the model and constraints, laying a scientific foundation for subsequent closed-loop verification and process card generation.

[0053] Step S140: Adaptively divide the injection speed into segments based on the shape of the simulated filling curve, and calculate the speed value of each segment based on the injection time setting parameter.

[0054] In this step, the flow characteristics of the melt in the cavity can be intelligently determined by analyzing the shape of the simulated filling curve output by CAE simulation, so as to realize the adaptive division of the injection speed segment.

[0055] Specifically, refer to Figure 4 , Figure 4 This is a flowchart of the steps for adaptively dividing the number of injection speed segments based on the shape of the simulated filling curve, provided in an embodiment of this application, including but not limited to steps S410 to S420.

[0056] Step S410: If the simulated filling curve is a single-peak curve, the injection speed segment is adaptively divided into 3 segments. Step S420: If the simulated filling curve is a bimodal or multimodal curve, the injection speed segment is adaptively divided into 4 segments.

[0057] In this embodiment, when the simulated filling curve is unimodal, the system automatically divides it into three segments (e.g., low-speed gate filling, medium-speed filling body, and high-speed holding pressure) to balance flow stability and efficiency. When the simulated filling curve exhibits a bimodal or multimodal shape, the system identifies the risk of melt merging or backflow and dynamically expands it into four segments, precisely controlling the injection speed in key areas to eliminate cavitation and weld lines. Subsequently, based on user-defined filling time parameters (e.g., 0.05-1.5 seconds for 25% filling time), combined with mold geometry and material flowability, the system automatically calculates the speed value for each segment using an algorithm, ensuring that the speed at any stage does not exceed 80% of the maximum allowable speed of the injection molding machine. This ensures molding quality while achieving the scientific and automated generation of process parameters.

[0058] Step S150: Based on the target screw stroke, target back pressure value, number of injection speed segments, and speed values ​​of each segment, and in conjunction with actual production conditions, generate a trial molding process card. The trial molding process card is used to guide the trial molding operation.

[0059] In this step, the trial molding process card is the core output that transforms CAE simulation and closed-loop dynamic compensation results into executable production instructions. It integrates verified optimal process parameters to ensure accurate and efficient trial molding operations. Based on the calibrated target screw stroke and back pressure value, combined with the number of injection speed segments (3 segments for single peaks, 4 segments for double / multi-peaks) adaptively divided by the CAE filling curve and the precise speed values ​​of each segment, the system automatically integrates actual production conditions such as equipment capacity boundaries, material characteristics, and mold status to generate a structured and standardized process card file. This process card contains complete parameters such as injection pressure, holding time, and cooling cycle, and can be directly sent to the injection molding machine control system to achieve the intelligent trial molding goal of "one-time mold installation, one-time calibration, and one-time qualification".

[0060] In some embodiments, refer to Figure 5 , Figure 5 This is a flowchart of the steps to be executed after generating a trial molding process card based on the target screw stroke, target back pressure value, number of injection speed segments and speed values ​​of each segment, combined with actual production conditions, according to an embodiment of this application. These steps include, but are not limited to, steps S510 to S520.

[0061] Step S510: When the injection molding machine is detected to be replaced, the current injection molding machine model is automatically identified and the safe operating boundary of the current injection molding machine is retrieved from the preset parameter table. Step S520: Automatically check the process parameters in the generated trial molding process card against the safe operating boundary, and automatically reduce and adjust the process parameters when they exceed the corresponding safe operating boundary so that the process parameters do not exceed the corresponding safe operating boundary.

[0062] In this embodiment, when a change of injection molding machine is detected, the system can automatically identify the new machine model and retrieve its specific safe operating boundaries (such as maximum injection pressure, maximum injection speed, and allowable back pressure range) from a preset equipment parameter library. Then, all key parameters in the generated trial molding process card (including target screw stroke, back pressure value, and injection speed at each stage) are automatically checked against the boundary conditions of the new machine. If any parameter exceeds the safety threshold, the system will immediately activate an intelligent derating mechanism, performing seamless optimization adjustments to the parameters according to preset priorities (such as prioritizing speed reduction followed by pressure adjustment). This ensures that all output commands operate within the physical safety range of the equipment, thereby completely eliminating the risk of parameter exceeding limits due to equipment differences without interrupting the trial molding process, and achieving seamless migration and safe execution of the process card between different machines.

[0063] In some embodiments, refer to Figure 6 , Figure 6 This is a second flowchart of an embodiment of the injection molding closed-loop control method provided in this application, including but not limited to steps S610 to S670.

[0064] Step S610: Obtain the theoretical filling parameters output by the computer-aided engineering simulation. The theoretical filling parameters include the theoretical melt density and the simulation filling curve. Step S620: Calculate the theoretical target product weight based on the theoretical melt density and mold cavity volume; Step S630: Call the pre-trained linear correlation model based on historical data, and back-calculate and optimize the target screw stroke and target back pressure value based on the theoretical target product weight and theoretical melt density. The linear correlation model is used to characterize the linear correlation between product weight, screw stroke and back pressure value. Step S640: Adaptively divide the number of injection speed segments according to the shape of the simulated filling curve, and calculate the speed value of each segment based on the injection time setting parameter. Step S650: Based on the target screw stroke, target back pressure value, number of injection speed segments, and speed values ​​of each segment, and combined with actual production conditions, generate a trial molding process card. The trial molding process card is used to guide the trial molding operation. Step S660: When the melt material is changed, load the theoretical melt density corresponding to the changed material from the material database; Step S670: Re-trigger CAE simulation based on material replacement to generate simulation filling curve corresponding to the replacement material, and return to the step of calculating the theoretical target product weight based on theoretical melt density and mold cavity volume, so as to generate trial molding process card corresponding to the replacement material based on theoretical melt density and simulation filling curve corresponding to the replacement material.

[0065] In this embodiment, by constructing an intelligent closed loop of "simulation-driven—dynamic verification—adaptive generation," a precise connection between CAE theory and industrial practice can be achieved. Specifically, the system first acquires the theoretical melt density and simulated filling curve output by CAE, and calculates the theoretical target product weight accordingly. Then, it calls a linear correlation model verified by historical data to deduce the initial screw stroke and initial back pressure value. By obtaining the actual product weight through the first injection, the system calculates the density and compares it with the theoretical density. If the error exceeds a preset threshold (e.g., 13%), it automatically iterates and corrects until the error is less than the preset threshold (e.g., 13%). At the same time, the system adaptively divides the injection speed segments according to the filling curve shape (single-peak 3 segments, double-peak / multi-peak 4 segments) and calculates the speed of each segment in conjunction with the set filling time. Finally, the calibrated screw stroke, back pressure value, and speed curve are integrated to generate a trial molding process card that can be directly executed. When materials are changed, the system automatically loads the theoretical density of the new material from the material database, triggers CAE re-simulation to obtain a matching simulation filling curve, and automatically backtracks to the target weight calculation stage (i.e., returns to step S620) to regenerate a complete process card adapted to the new material. No manual intervention is required throughout the process, ensuring that the process parameters evolve dynamically with the material properties, and truly achieving the intelligent trial molding goal of "one-time mold assembly, multiple material adaptation, and one-time qualification".

[0066] Reference Figure 7 , Figure 7 This is a structural block diagram of an injection molding trial mold closed-loop control system provided in one embodiment of this application. This application also provides an injection molding trial mold closed-loop control system 700, which may include: Material quantity compensation algorithm module 710 is used to perform the following steps: Obtain the theoretical filling parameters output by the computer-aided engineering simulation. The theoretical filling parameters include the theoretical melt density and the simulation filling curve. Calculate the theoretical target product weight based on the theoretical melt density and mold cavity volume; The linear correlation model, which is pre-trained using historical data, is invoked to back-calculate and optimize the target screw stroke and target back pressure value based on the theoretical target product weight and theoretical melt density. The linear correlation model is used to characterize the linear correlation between product weight, screw stroke, and back pressure value. The screw curve optimization module 720 is used to adaptively divide the number of injection speed segments according to the shape of the simulated filling curve, and calculate the speed value of each segment based on the injection setting time parameter. The process card generation module 730 is used to generate trial molding process cards based on the target screw stroke, target back pressure value, number of injection speed segments, and speed values ​​of each segment, combined with actual production conditions. The trial molding process cards are used to guide trial molding operations.

[0067] In this embodiment, the injection molding trial mold closed-loop control system 700 constructs a full-link intelligent closed loop from CAE simulation to industrial execution through the coordinated operation of three core modules (including the material quantity compensation algorithm module 710, the screw curve optimization module 720, and the process card generation module 730). Specifically, the material quantity compensation algorithm module 710 first receives the theoretical melt density and simulated filling curve output from the CAE, calculates the theoretical target product weight based on the cavity volume, and calls a linear correlation model validated by historical data to back-calculate and optimize the target screw stroke and back pressure value under the physical constraints of the equipment, achieving a precise mapping from theoretical density to executable mechanical parameters. The screw curve optimization module 720 intelligently divides the injection speed range into 3 or 4 segments based on the shape of the simulated filling curve (single-peak, double-peak, or multi-peak), and automatically calculates the speed value of each segment in conjunction with the user-set filling time parameters, ensuring dynamic matching between flow control and melt behavior. The process card generation module 730 ultimately integrates the calibrated screw stroke, back pressure value, and adaptive speed curve, and combines the actual conditions such as equipment capacity and material characteristics at the production site to generate a structured trial molding process card that can be directly sent to the injection molding machine. The entire process requires no manual intervention, realizing an integrated scientific trial molding process of "simulation input - intelligent calculation - automatic output", and completely bridging the data gap between theoretical analysis and actual production.

[0068] In some embodiments, continue to refer to Figure 7 The injection molding trial mold closed-loop control system 700 may also include a data acquisition module 740. The data acquisition module 740 is used to acquire the screw stroke, back pressure value, and actual weight of the product of the injection molding machine.

[0069] In this embodiment, the data acquisition module 740 is the core sensing layer of the closed-loop control system, realizing the "perception-decision-execution" closed loop. It collects three key physical quantities (screw stroke, back pressure value, and actual product weight) in real time and without human intervention through a high-precision industrial sensor network. The screw stroke can be accurately measured by a magnetostrictive displacement sensor mounted on the piston rod of the injection cylinder. Its non-contact design ensures long-term stable operation in the high-temperature, high-humidity, and oily injection molding environment. The back pressure value can be collected in real time by a piezoresistive pressure sensor directly integrated into the injection cylinder or flow channel, directly converting the hydraulic system pressure into an electrical signal, ensuring accurate monitoring of the back pressure during the melt plasticizing stage. The actual product weight is automatically determined by a high-precision weighing sensor integrated below the demolding conveyor mechanism after demolding, enabling automatic weighing of the product the instant it is removed from the mold, without any manual intervention. All sensor data is transmitted in real time to the core control unit via industrial-grade communication protocols (such as EtherCAT or Modbus TCP) at a sampling frequency of no less than 1kHz. This provides a unique and reliable data source for the system's computational density and is the physical foundation for achieving dynamic verification and adaptive optimization in the "one-time trial pass" process.

[0070] In some embodiments, continue to refer to Figure 7 The injection molding trial mold closed-loop control system 700 may also include a human-machine interface terminal 750. The human-machine interface terminal 750 is used to display the process parameters of the trial mold process card and to display the segmented adjustment interface for the number of injection speed segments.

[0071] In this embodiment, the human-machine interface terminal 750 serves as the core of human-machine collaboration in the closed-loop control system. It is the only visual interface connecting the intelligent algorithm and on-site operation. Through a high-resolution touchscreen, it dynamically displays the complete trial molding process card parameters after closed-loop verification in real time, including the locked target screw stroke, back pressure value, injection pressure and holding time of each segment, and presents the CAE adaptively divided injection speed curve in a graphical manner. For example, a single-peak curve is clearly marked with three color blocks (such as low-speed gate filling, medium-speed filling, and high-speed holding pressure), while a double-peak / multi-peak curve is expanded to four segments. The speed value of each segment is accurately labeled with a numerical label. At the same time, it supports the operator to directly drag curve nodes on the terminal interface, fine-tune the speed of any segment, or input the target filling time (0.05–1.5 seconds). The system will respond instantly and recalculate the speed of each segment to ensure that the adjusted parameters are always within the equipment safety boundary. The 750 human-machine interaction terminal not only makes process parameters transparent and readable, but also empowers operators to make "human-machine collaborative fine-tuning" decisions based on the system's automatic convergence, upgrading "first-time trial molding qualification" from a purely automated process to a highly efficient production mode that combines intelligent decision-making with professional experience.

[0072] The preferred embodiments of the present application have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and substance of the embodiments of the present application shall be within the scope of the claims of the present application.

Claims

1. A method of closed loop control of injection molding test molds, characterized in that, The method comprises the following steps: obtaining theoretical filling parameters of computer-aided engineering simulation output, the theoretical filling parameters comprising theoretical melt density and a simulation filling curve; calculating a theoretical target product weight based on the theoretical melt density and a mold cavity volume; calling a linear correlation model trained in advance through historical data to inversely deduce and optimize a target screw stroke and a target back pressure value based on the theoretical target product weight and the theoretical melt density, the linear correlation model being used to represent a linear correlation relationship among product weight, screw stroke and back pressure value; self-adaptively dividing injection speed segments according to a shape of the simulation filling curve and calculating speed values of each segment based on an injection setting time parameter; generating a trial mold process card according to the target screw stroke, the target back pressure value, the injection speed segments and the speed values of each segment, the trial mold process card being used to guide trial mold operation.

2. The method of claim 1, wherein, The calling of the linear correlation model trained in advance through historical data to inversely deduce and optimize the target screw stroke and the target back pressure value based on the theoretical target product weight and the theoretical melt density comprises: calling the linear correlation model trained in advance through historical data to calculate a group of calculation data including screw stroke and back pressure value based on the theoretical target product weight and constraint conditions of a current working condition; obtaining an actual product weight obtained by executing injection under the condition of the calculation data; calculating a system calculation density based on the actual product weight and comparing the system calculation density with the theoretical melt density; if an error between the system calculation density and the theoretical melt density is greater than a set threshold, calling the linear correlation model to recalculate based on the theoretical target product weight and the constraint conditions of the current working condition and updating the calculation data until the error between the system calculation density and the theoretical melt density is less than or equal to the set threshold; taking the screw stroke and the back pressure value under the condition that the error between the system calculation density and the theoretical melt density is less than or equal to the set threshold as the target screw stroke and the target back pressure value.

3. The method of claim 2, wherein, The calling of the linear correlation model trained in advance through historical data to calculate a group of calculation data based on the theoretical target product weight and constraint conditions of a current working condition comprises: substituting the theoretical target product weight into the linear correlation model and establishing a solving equation; under the constraint conditions of the current working condition, selecting a group of optimal solutions from all combinations of feasible solutions according to a preset optimization target to obtain a group of calculation data.

4. The method of claim 2, wherein, The calculation of the system calculation density based on the actual product weight comprises: calculating the system calculation density based on the actual product weight, corresponding screw stroke and screw cross-sectional area.

5. The method of claim 1, wherein, The self-adaptively dividing of injection speed segments according to the shape of the simulation filling curve comprises: if the simulation filling curve is a unimodal curve, self-adaptively dividing the injection speed segments into three segments; if the simulation filling curve is a bimodal or multimodal curve, self-adaptively dividing the injection speed segments into four segments.

6. The method of claim 1, wherein, After generating the trial mold process card according to the target screw stroke, the target back pressure value, the number of injection speed segments, and the speed value of each segment in combination with the actual production conditions, the method further comprises: When the injection molding machine is detected to be replaced, the current injection molding machine model is automatically identified, and the safe operation boundary of the current injection molding machine is called from the preset parameter table; The generated process parameters in the trial mold process card are automatically checked against the safe operation boundary, and when the process parameters exceed the corresponding safe operation boundary, automatic reduction adjustment is performed so that the process parameters do not exceed the corresponding safe operation boundary.

7. The method of claim 1, wherein, After generating the trial mold process card according to the target screw stroke, the target back pressure value, the number of injection speed segments, and the speed value of each segment in combination with the actual production conditions, the method further comprises: When the melt material is replaced, the theoretical melt density corresponding to the replaced material is loaded from the material database; Based on the replaced material, the CAE simulation is retriggered to generate a simulation filling curve corresponding to the replaced material; The step of calculating the theoretical target product weight based on the theoretical melt density and the mold cavity volume is returned to generate a trial mold process card corresponding to the replaced material based on the theoretical melt density corresponding to the replaced material and the simulation filling curve corresponding to the replaced material.

8. An injection molding test mold closed loop control system, characterized by, Comprise: A material quantity compensation algorithm module for performing the following steps: Obtain theoretical filling parameters output by computer-aided engineering simulation, the theoretical filling parameters including a theoretical melt density and a simulation filling curve; Calculate a theoretical target product weight based on the theoretical melt density and a mold cavity volume; Call a linear correlation model trained in advance through historical data, and inversely deduce and optimize the target screw stroke and the target back pressure value based on the theoretical target product weight and the theoretical melt density, the linear correlation model being used to represent a linear correlation relationship between the product weight, the screw stroke, and the back pressure value; A screw curve optimization module for adaptively dividing the number of injection speed segments according to the shape of the simulation filling curve, and calculating the speed value of each segment based on an injection setting time parameter; A process card generation module for generating a trial mold process card according to the target screw stroke, the target back pressure value, the number of injection speed segments, and the speed value of each segment in combination with the actual production conditions, the trial mold process card being used to guide trial mold operation.

9. The system of claim 8, wherein, The system further comprises: A data acquisition module for acquiring the screw stroke, the back pressure value, and the actual product weight of the injection molding machine.

10. The system of claim 8, wherein, The system further comprises: A human-computer interaction terminal for displaying the process parameters of the trial mold process card and showing a segmented adjustment interface of the number of injection speed segments.

Images