Plastic part injection mold regulation and control system and method based on complex curved surface and thin-wall structure

Abstract

The invention relates to the technical field of injection molding, and discloses a plastic part injection mold regulation and control system and method based on a complex curved surface and a thin-wall structure, and the plastic part injection mold regulation and control system based on the complex curved surface and the thin-wall structure comprises a mold modeling module, a temperature modeling module, a control modeling module, a control optimization module and a control execution module; the plastic part injection mold regulation and control method based on the complex curved surface and the thin-wall structure comprises the steps that a mold three-dimensional model and thermal control area mapping are built, a heat conduction model is built, modal dimensionality reduction is conducted, a target temperature field and a performance function are combined, optimal temperature regulation and control are achieved through a predictive control algorithm, real-time feedback is matched, and a heating and cooling unit is driven. And dynamic closed-loop control of the mold thermal field is realized. According to the method, thermal control precision and energy consumption efficiency are optimized through Galerkin modal dimension reduction, model predictive control and a weighted performance function, global closed-loop temperature control is achieved by combining mapping of a three-dimensional area and a heating and cooling unit of the mold, and system safety and stability are enhanced.

Description

Technical Field

[0001] This invention relates to the field of injection molding technology, specifically to a control system and method for injection molds of plastic parts based on complex curved surfaces and thin-walled structures. Background Technology

[0002] In the precision injection molding process of plastic parts with complex curved surfaces and thin-walled structures, mold temperature control directly affects the molding quality and dimensional stability of the product. Existing temperature control methods mostly rely on single-point temperature measurement combined with heating elements or cooling channels for regional adjustment, with control strategies primarily based on empirical adjustment or traditional PID control. While this approach is structurally simple, it struggles to construct effective mathematical models for global control when dealing with the multi-spatial coupling and dynamic disturbances of the heat conduction process within the mold. This results in slow response to temperature changes, and localized overheating or insufficient cooling, ultimately affecting product consistency and production stability.

[0003] Furthermore, some current research attempts to introduce the finite element method (FEM) for mold temperature field simulation to guide mold temperature setting or preheating operations. However, these numerical solutions typically involve large computational loads and slow response times, making the models themselves difficult to apply to real-time iterations in actual control systems. Even in systems that employ preset temperature control curves, there is generally a lack of effective coupling with real-time measurement feedback, making it impossible to adjust control strategies during rapid switching of operating conditions. This results in poor system adaptability to sudden heat source disturbances, and limited control precision and flexibility.

[0004] More importantly, most existing control systems lack the ability to modally identify and map the overall temperature field of the mold, and temperature control only makes local adjustments to the area near the sensor point. This one-sided control method ignores the thermal coupling relationship between the complex thermal structures of the mold, and often fails to solve the problem of overall temperature field imbalance in the mold. Especially in mold designs involving complex geometries or non-homogeneous materials during injection molding, this technical approach is not robust enough, and the control strategy is difficult to adapt to changing process requirements. Therefore, this invention proposes a plastic injection mold control system and method based on complex curved surfaces and thin-walled structures to address the shortcomings of existing technologies. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a plastic injection mold control system and method based on complex curved surfaces and thin-walled structures, which solves the problems of slow response, uneven adjustment, and lack of global dynamic closed-loop control capability in mold thermal field control.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a plastic injection mold control system based on complex curved surfaces and thin-walled structures, comprising:

[0007] The mold modeling module is used to construct the three-dimensional geometry of precision injection molds and divide them into multiple independent thermal control areas for the temperature modeling module to establish a heat conduction model.

[0008] The temperature modeling module is used to establish a heat conduction model of the mold and perform physical modeling, providing thermophysical equation inputs for the control modeling module.

[0009] The control modeling module is used to perform finite-dimensional modeling transformation of the heat conduction model, providing a state-space model for the control optimization module;

[0010] The control optimization module is used to construct a predictive controller for the dimensionality-reduced thermal field state system and output control commands to the control execution module.

[0011] The control execution module is used to drive the mold's distributed heating and cooling unit and form a closed-loop control, and to feed back the current temperature of the mold to the control optimization module.

[0012] This invention also provides a method for controlling injection molds for plastic parts based on complex curved surfaces and thin-walled structures, comprising the following steps:

[0013] S1. Construct a three-dimensional geometric model of the precision injection mold and divide it into multiple thermal control zones, and establish the spatial mapping relationship between each zone and the heating and cooling unit;

[0014] S2. Establish a physical model of heat conduction in precision injection molds, set the thermal conductivity, initial temperature distribution and boundary heat transfer conditions, and form unsteady-state heat conduction equations.

[0015] S3. Select spatial orthogonal basis functions, perform Galerkin mode dimensionality reduction on the heat conduction equation, construct a finite-dimensional state-space control model, and obtain the modal state matrix, input matrix and disturbance term expression;

[0016] S4. Map the target temperature field distribution to the modal space and generate the target modal state vector as a control reference;

[0017] S5. Construct a performance function, which includes a state deviation term, a control input term, and a control change term, and introduce a corresponding weighting matrix;

[0018] S6. Based on the state-space model and performance function, a predictive control algorithm is used to solve for the optimal control input sequence online, and the control input corresponding to the current control step is extracted.

[0019] S7. Collect multi-point temperature data of the mold and map it to the modal state space as the current state input controller of the system;

[0020] S8. Based on the control input obtained from the solution, drive the distributed heating and cooling units to regulate the temperature, thereby realizing dynamic closed-loop control of the thermal field of the precision injection mold.

[0021] This invention provides a system and method for controlling injection molds for plastic parts based on complex curved surfaces and thin-walled structures. It has the following beneficial effects:

[0022] 1. This invention employs a modal dimensionality reduction method based on Galerkin projection to transform a complex unsteady heat conduction model into a controllable finite-dimensional state-space model, achieving the technical effects of reducing computational complexity and improving the real-time performance of control modeling. Compared to the existing method of directly solving partial differential models numerically, this avoids the problem of high computational resource consumption and solves the problems of poor real-time performance and difficulty in online control.

[0023] 2. This invention introduces a model predictive control structure and constructs a weighted performance function that includes state deviation, control energy consumption, and control variation terms. This achieves a dynamic balance between thermal control accuracy and energy efficiency. Compared to traditional single PID temperature control strategies, it can proactively predict temperature change trends, overcoming the shortcomings of control response lag and lack of global optimization capabilities.

[0024] 3. This invention establishes a mapping relationship between the three-dimensional region of the mold and the distributed heating and cooling units. Combined with multi-point temperature acquisition and modal state feedback mechanisms, it achieves system-level closed-loop temperature control. Unlike conventional systems that only control temperature locally, this invention can adjust the entire mold's thermal field globally, significantly improving the problems of uneven heat distribution and large molding errors.

[0025] 4. This invention maps the target thermal field distribution to modal space for control reference tracking, and combines signal limiting and jump constraint strategies to enhance the safety of control execution and system stability. Compared with existing systems that do not consider the physical boundaries of actuators, this invention effectively avoids potential risks such as control overshoot and hardware overload, making it suitable for long-term stable operation in industrial settings. Attached Figure Description

[0026] Figure 1 This is a system architecture diagram of the present invention;

[0027] Figure 2 This is a flowchart of the method of the present invention. Detailed Implementation

[0028] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] Please see Figure 1 This invention provides a plastic injection mold control system based on complex curved surfaces and thin-walled structures, comprising:

[0030] The mold modeling module is used to construct the three-dimensional geometry of precision injection molds and divide them into multiple independent thermal control areas for the temperature modeling module to establish a heat conduction model.

[0031] The mold modeling module in this invention is mainly used to construct the three-dimensional geometry of a precision injection mold and divide the mold into multiple independent thermal control zones for use by the subsequent temperature modeling module. The output of this module provides fundamental data for the thermal control of the entire system and important input information for the subsequent control optimization module. Through precise geometric modeling and reasonable thermal control zone division, the mold modeling module ensures the efficiency and accuracy of subsequent heat conduction simulation and control processes.

[0032] In this embodiment, the mold modeling module consists of three sub-modules: a 3D modeling unit, a region division unit, and a control interface calibration unit. The functions of each sub-module are described in detail below.

[0033] In this embodiment, the main task of the 3D modeling unit is to construct a parametric 3D geometric model of a precision injection mold. Typically, injection molds used for molding complex curved surfaces and thin-walled plastic parts have complex geometries and include multiple injection channels, cooling channels, and heating zones. The 3D modeling unit utilizes computer-aided design (CAD) technology to generate an accurate 3D geometric model based on input design parameters (such as the mold's external dimensions, structural shape, and functional requirements). This model reflects the spatial structure of the mold and provides the necessary geometric information for subsequent heat conduction analysis and temperature control modeling.

[0034] During the modeling process, 3D modeling units typically employ parametric modeling techniques, which use a set of parameters (such as length, width, and height) to define the geometry of the mold. These parameters can be flexibly adjusted to adapt to the design of different headphone mold models. This approach ensures a high degree of consistency between the mold's structural shape and functional requirements.

[0035] Specifically, the 3D modeling unit completes the modeling process through the following steps:

[0036] Input design parameters (such as mold size, structure, etc.).

[0037] Based on these parameters, a three-dimensional geometric model of the mold is constructed using CAD software.

[0038] For complex parts, a segmented modeling method is adopted, which combines different sub-models to form a complete mold geometry.

[0039] In one possible implementation, the 3D modeling unit may also generate the distribution of injection channels, cooling channels, and heating zones based on the functional requirements of the mold. This data will directly affect the subsequent division of thermal control zones and heat conduction modeling.

[0040] The task of the zone division unit is to divide the three-dimensional geometry of the mold into multiple independent thermal control zones. Each thermal control zone requires different temperature control precision based on the thermal conduction characteristics of different parts of the mold. Therefore, reasonable zone division is crucial for subsequent thermal control.

[0041] Generally, the thermal control zone is divided based on the following factors:

[0042] The functional areas of a mold, such as the injection area, cooling area, and heat source area, are divided according to the working requirements of different areas.

[0043] Thermal conductivity characteristics: The thermal conductivity, temperature response and other characteristics vary greatly in different regions, and it is necessary to distinguish them based on these characteristics.

[0044] Control precision requirements: Some areas (such as the injection molding area of ​​the mold) have high requirements for temperature control, while other areas (such as the cooling area) have lower requirements for control precision.

[0045] Specifically, the regional division units may be divided using the following strategies:

[0046] Uniform division: Divide the entire mold into several thermal control zones evenly. This method is simple and effective, but it may not be able to optimize the heat conduction characteristics of different zones.

[0047] Non-uniform division based on heat conduction efficiency: Fine division is carried out on different areas of the mold surface based on temperature sensitivity and heat conduction path to ensure temperature control accuracy.

[0048] For example, in precision injection molds, the injection zone of the mold usually requires high temperature control accuracy, so it can be divided into smaller thermal control areas, while the cooling zone and support structure outside the mold can be divided into larger thermal control areas.

[0049] The task of the control interface calibration unit is to ensure the accurate spatial mapping between the thermal control zone and the heating and cooling control channels. This process is crucial for achieving precise temperature control because the control signals from the heating and cooling units must be applied accurately to each thermal control zone.

[0050] Specifically, the control interface calibration unit establishes a thermodynamic mapping relationship between each thermal control zone and control channel based on the mold's three-dimensional geometric model, using numerical simulations, experimental data, or known heat transfer laws. The calibration process considers not only the location of the thermal control zones but also the thermal response characteristics of each zone. For example, in some zones, heat transfer efficiency is higher; the calibration process takes this into account to determine the control efficiency of the heating and cooling channels.

[0051] By using the calibrated mapping relationship, the control optimization module can more accurately adjust the temperature of each thermal control zone and achieve optimized control of the system.

[0052] The output of the mold modeling module is the input of the temperature modeling module. Specifically, the 3D geometric model and thermal control zone information generated by the mold modeling module provide the necessary geometric data for the temperature modeling module, enabling the latter to establish an accurate heat conduction model and calculate the temperature field of the mold. Since different thermal control zones have different thermal responses, the accuracy of zone division and interface calibration directly affects the accuracy of subsequent temperature simulations.

[0053] In the mold modeling module, 3D geometric modeling and region partitioning are crucial factors affecting the accuracy of the heat transfer model. Specifically, the mold heat transfer model can be described by the following equations:

[0054] ;

[0055] in: Indicates the location Location, Time The temperature at that time; It is the gradient operator; It is a location Thermal conductivity at that point, expressed in W / (m·K); It is the heat source distribution per unit volume, with units of W / m². 3 .

[0056] This equation describes the heat propagation process inside the mold, taking into account the influence of thermal conductivity and external heat sources on the temperature distribution of the mold. By solving this equation through numerical simulation (such as the finite element method), the temperature distribution at various points inside the mold over time can be obtained.

[0057] To enhance the effectiveness of the technical solution, this invention can also employ a multi-level thermal control zone division strategy to further improve the accuracy of thermal control. For example, the injection zone of the mold can be further subdivided into multiple sub-regions, each connected to an independent heating and cooling unit, thereby achieving more precise temperature control. Furthermore, for each thermal control zone, heat source simulation can be introduced to more realistically reflect the impact of heat source changes on the mold temperature field during injection molding.

[0058] Through these optimizations, the mold modeling module can not only provide accurate geometric models and thermal control zone divisions, but also provide more detailed and reliable data support for subsequent heat conduction analysis and control optimization, thereby improving the overall temperature control effect of the system.

[0059] The temperature modeling module is used to establish a heat conduction model of the mold and perform physical modeling, providing thermophysical equation inputs for the control modeling module.

[0060] The temperature modeling module in this invention is responsible for providing the thermophysical equations as input for the subsequent control modeling module. This module's task is to establish a mathematical model of heat transfer during the injection molding process based on the three-dimensional geometry of the precision injection mold. Through accurate heat conduction modeling, the temperature modeling module provides the necessary thermophysical equations for the control modeling module, ensuring the efficiency and accuracy of the subsequent temperature control process. The detailed technical solution for the temperature modeling module is as follows.

[0061] The temperature modeling module mainly consists of three key sub-units: a heat conduction modeling unit, a parameter setting unit, and a boundary condition setting unit. Each unit provides a reliable thermophysical model for the entire control system through precise data processing and physical modeling.

[0062] In this embodiment, the heat conduction modeling unit is mainly responsible for establishing a thermophysical model based on the three-dimensional unsteady-state heat conduction law according to the geometry of the precision injection mold. Since different parts of the mold will have different thermal responses during temperature control, it is necessary to accurately model the heat conduction process of the entire mold.

[0063] Generally, heat conduction problems can be described by the heat conduction equation. For unsteady-state heat conduction problems, the commonly used equation is:

[0064] ;

[0065] in: For position Location, Time Temperature at that time (in K); This is the gradient operator, representing the rate of change of temperature in space; For position Thermal conductivity at a given point (measured in W / (m·K)) is a physical quantity that describes the heat transfer capacity of a material. Heat source distribution per unit volume (unit: W / m²) 3 This describes the heat released by the material during the injection molding process; This is the divergence operator, representing the degree of heat diffusion.

[0066] Specifically, the first term in the heat conduction equation The process of heat diffusion through the mold is described, while the second term... It describes the temperature change caused by a heat source. This equation is the standard description of unsteady heat conduction in three-dimensional space, and therefore needs to be solved by numerical methods (such as the finite element method, finite difference method, etc.).

[0067] In one possible implementation, the equation can be discretized using the finite element method, and the temperature distribution at each location within the mold can be obtained through iterative solution.

[0068] The parameter setting unit in the temperature model block provides the necessary physical parameters for the heat conduction modeling unit to accurately simulate the heat conduction process of the mold. Accurate physical parameters are crucial for simulating the temperature field; incorrect parameters may lead to instability or ineffective operation of the control system.

[0069] Specifically, the key parameters provided by the parameter setting unit include:

[0070] thermal conductivity Thermal conductivity is a parameter describing a material's ability to transfer heat, and it typically varies with temperature. Different materials have different values ​​for thermal conductivity. For example, aluminum alloys have high thermal conductivity, while steel has low thermal conductivity. In a specific practical application, thermal conductivity... Data can be obtained through experimental measurements or from materials databases.

[0071] Volumetric heat source Volumetric heat sources refer to the heat generated in the mold during the injection molding process. This heat source typically varies over time. In the headphone injection molding process, the heat source primarily comes from the contact between the molten material and the mold. The distribution of volumetric heat sources can be obtained by simulating melt flow or directly from experimental data.

[0072] Heat transfer boundary conditions: The heat exchange between the mold surface and the external environment has a significant impact on the overall temperature field of the mold. The parameter setting unit sets the boundary exchange coefficient based on factors such as the mold surface material, the ambient temperature, and the presence of a cooling system. Specifically, the boundary heat transfer conditions can be described by the following formula:

[0073] ;

[0074] in: Heat flux density at the boundary (in W / m³) 2 ); The heat exchange coefficient (unit: W / (m²·K)); Temperature at the mold surface (in K); The temperature of the external environment (in K).

[0075] As an alternative, the boundary heat exchange coefficient The settings can be made based on experimental data or numerical simulations to ensure that the temperature simulation accurately reflects the heat exchange during the cooling or heating process.

[0076] The boundary condition setting unit is mainly responsible for setting boundary conditions based on the heat exchange conditions of the external environment of the mold. Typically, heat exchange between the mold surface and the external environment includes three main forms: convection, radiation, and conduction. These factors work together to affect the temperature distribution of the mold.

[0077] In one possible implementation, the mold surface may come into contact with different cooling media (such as water or air) or with heating elements (such as electric heaters or heating films). Therefore, the boundary condition setting unit sets appropriate boundary heat flux conditions according to different environmental conditions.

[0078] For example, if the mold surface is in contact with water cooling channels, a higher heat exchange coefficient may be required. However, if the mold is in the air, the heat exchange coefficient... The results will be lower. Based on experimental or simulation results, the boundary conditions can be dynamically adjusted to adapt to different cooling and heating environments.

[0079] The output of the temperature modeling module directly affects the input of the control modeling module. The mold modeling module provides the mold's three-dimensional geometry and thermal control zone division, which provides necessary data support for the temperature modeling module. In the temperature modeling module, the heat conduction modeling unit simulates the mold's temperature field based on the three-dimensional geometric model, combined with physical parameter settings and boundary conditions.

[0080] The control modeling module receives the thermophysical equations output by the temperature modeling module, thereby transforming the heat conduction problem into a state-space model suitable for control optimization. In this way, the temperature modeling module and the control modeling module achieve seamless integration.

[0081] To improve the accuracy of modeling and the overall performance of the system, the temperature modeling module in this invention can also incorporate other supplementary technical details. For example, a temperature-dependent material model can be introduced, taking into account the changes in the thermal conductivity and specific heat capacity of the material with temperature. Furthermore, considering the potential non-uniform heat source distribution within the mold, the temperature modeling module can also perform more refined modeling of the heat source distribution to ensure more accurate simulation results.

[0082] In some embodiments, the temperature modeling module can also employ multi-level, multi-scale modeling methods to simulate the thermal effects that may occur in the mold during more complex injection molding processes. These technical details help to further improve temperature control accuracy, especially ensuring that the system can maintain stable temperature control even under complex operating conditions during production.

[0083] The control modeling module is used to perform finite-dimensional modeling transformation of the heat conduction model, providing a state-space model for the control optimization module;

[0084] In the precision injection mold-based control system described in this invention, the temperature modeling module has already achieved unsteady-state physical modeling of the mold's heat conduction behavior. However, to achieve controllability and predictability of the thermal control system, this partial differential model needs to be converted into a finite-dimensional ordinary differential state-space model suitable for dynamic control solutions. Based on this, this invention further includes a control modeling module to complete the mapping process from physical modeling to control modeling, realizing dimensionality reduction modeling and structured transformation of the mold's heat conduction system.

[0085] This module serves as a logical bridge between the temperature modeling module and the control optimization module. It constructs a finite-dimensional state-space model using the Galerkin method, providing a mathematical foundation and input support for optimizing control strategies. This module comprises a basis function construction unit, a Galerkin projection unit, and a state coefficient generation unit. These three sub-units work collaboratively to complete the system transformation from a continuous space model to a discrete state model.

[0086] In practical applications, basis function building units are used to establish a set of spatially orthogonal basis functions within the three-dimensional spatial region of a computer-generated injection mold, supporting the spatial unfolding of Galerkin projection operations. Typically, the mold region... It is considered a finite, closed region with boundary conditions.

[0087] Specifically, a set of spatially orthogonal basis functions is defined. They satisfy the following orthogonality conditions:

[0088] ;

[0089] in: For the first A spatial basis function is defined in the three-dimensional region of the mold. ; It is a spatial coordinate vector. ; For Kronecker notation, when The value is 1 if it is true, and 0 otherwise. To truncate the dimension, control the upper limit of the dimension of the state model.

[0090] Alternatively, the basis functions may be polynomial functions (such as Legendre and Chebyshev functions) or piecewise linear basis functions constructed under the finite element method. In some embodiments, B-spline functions may also be introduced to improve local approximation capabilities.

[0091] The main function of the Galerkin projection unit in this example is to project the three-dimensional heat conduction differential equation model output by the temperature modeling module into space using orthogonal basis functions, thereby obtaining a finite-dimensional ordinary differential state space form.

[0092] Set mold temperature field The following heat conduction governing equations must be satisfied:

[0093] ;

[0094] in: For position At the moment Temperature, in Kelvin; For position Thermal conductivity of the material, expressed in W / (m·K); This is the heat source term per unit volume, with units of W / m³. 3 ; To control the input spatial action function (such as the heat source distribution of a heater or cooling device), the unit is 1 / m. 3 ; For a moment The control input, in W; It is the divergence operator; This is the gradient operator.

[0095] Modal expansion of the temperature field yields:

[0096] ;

[0097] in, Temperature field in space and time Temperature distribution below; For the first The modal coefficient represents the th modal coefficient. Each mode in time The strength; For the first A spatial basis function, usually a set of orthogonal functions; is the modality number, and the number of spatial basis functions retained.

[0098] After substituting into the original partial differential equation, for each basis function Projecting the test function, we obtain the following modal differential system:

[0099] ;

[0100] in, For the first Time derivatives of each modal coefficient; The elements in the state matrix represent modes. For modes The coupling effect; The elements in the input matrix represent control inputs. For modes The impact; For the first Each controller in time The input value; For the first External disturbances experienced by each mode; To control the number of input channels.

[0101] Further expressed as the state-space equation in vector form:

[0102] ;

[0103] in, This is the modal state vector, in units of K; The state matrix; The input matrix is ​​the projection of the controller into the modal space. To control the input vector; Let be the perturbation vector, and its elements are defined as follows:

[0104] ;

[0105] ;

[0106] To control the input vector;

[0107] Let the perturbation vector be defined as:

[0108] ;

[0109] in, In position ,time External thermal disturbances on the surface;

[0110] The state-space model generation unit is responsible for generating all the coefficients of the final state-space model used for control optimization, including the state matrix, based on the results of the Galerkin projection. Control input matrix and disturbance terms .

[0111] In general, this unit will use numerical integration or analytical integration to calculate the coefficients of the above integral form. As an alternative, Gaussian integration or finite element interpolation can be used to solve for the values ​​of each coefficient matrix.

[0112] In some specific implementations, a state output matrix can be introduced. With observation output This can be expanded into a standard control system structure.

[0113] ;

[0114] in: This is the output matrix, used to convert state variables into physical measurements; This is the output quantity, such as the temperature value collected by the sensor.

[0115] This state-space model will serve as the predictive model input for subsequent control optimization modules (such as Model Predictive Controller, MPC), supporting the construction and optimization of the control objective function.

[0116] In an alternative implementation, the state coefficient generation unit can also construct a nonlinear state matrix, i.e., a state matrix, based on a temperature-dependent model. Can vary with mode vector Dynamic changes constitute the following nonlinear system:

[0117] ;

[0118] In this implementation, the system control can adapt to the nonlinear characteristics of thermal conduction in different temperature zones of the mold, enhancing the robustness of the control strategy. Furthermore, this module supports spatial controllability analysis and modal observer design, such as employing the Model Reduction method to retain dominant modes and discard high-frequency modes, thereby further reducing model computational complexity and improving online control performance.

[0119] The control optimization module is used to construct a predictive controller for the dimensionality-reduced thermal field state system and output control commands to the control execution module.

[0120] In the precision injection mold-based control system described in this invention, the control optimization module plays a crucial role in transforming the heat conduction state-space model into optimal control input commands. The aforementioned control modeling module has already reduced the dimensionality of the unsteady heat conduction partial differential equation model to a finite-dimensional state-space form based on the Galerkin method, thereby providing a computable dynamic model input for the predictive controller. Based on this state prediction model, the control optimization module constructs a performance objective function and calculates the optimal control input through an optimized solution strategy to achieve dynamic and precise control of the temperature distribution inside the mold.

[0121] In this embodiment, the control optimization module includes a modal target setting unit, a performance function construction unit, and an optimal control solution unit, which are used to respectively complete the target state projection, performance index definition, and real-time acquisition of the optimal control quantity. The implementation of this module ensures that the thermal control system achieves rapid response and smooth adjustment of temperature distribution under constrained conditions, adapting to the actual needs of industrial injection mold control scenarios.

[0122] In this experimental example, the simulated target setting unit receives the target temperature distribution set by the user or the system. And through the aforementioned set of basis functions The orthogonal projection generates the target modal state vector. . Specifically:

[0123] ;

[0124] in: Define the three-dimensional domain of the mold; The target thermal field temperature function is expressed in K. For the first A spatial orthogonal basis function; For the target modal coefficients, corresponding to the th Modal dimension; The target modal state vector is expressed in units of K.

[0125] Generally, the target thermal field can be preset based on the mold structure characteristics, injection molding material properties, or process experience data, or it can be provided by a simulation thermodynamic model.

[0126] In this experimental example, the performance function construction unit is used to construct a weighted control performance index function within the rolling optimization control framework. This serves as the objective function for subsequent optimization. This function comprehensively considers the deviation between the temperature distribution and the target value, the amplitude limitation of the control input itself, and the rate of change of the control input. Its construction is as follows:

[0127] ;

[0128] in: This refers to the overall performance index of predictive control, without units. The prediction time domain length is the number of steps the system predicts into the future, expressed in units of zero. For the first The modal state vector for each prediction step, in units of... ; The target modal state vector, in units of ; For the first The control input vector for each prediction step is expressed in W. This is the control input vector for the previous prediction step, in units of W; The modal error weighting matrix is ​​positive definite or positive semi-definite. The input weighting matrix is ​​positive definite. To control the rate of change weighting matrix, it can be positive definite or positive semi-definite.

[0129] The first term is the modal state error penalty term, which reflects the degree of deviation between the current predicted state and the target thermal field; the second term controls energy consumption by limiting the actual power through penalty input amplitude; and the third term is used to suppress drastic changes in the control signal and improve the dynamic stability of the system.

[0130] As an option, matrix It can be manually set, or it can be adaptively adjusted through offline training or reinforcement learning to meet the thermal control requirements of different operating conditions.

[0131] In this example, the optimal control solution unit receives the performance function. Using a finite-dimensional prediction model, a rolling optimization problem is constructed, and the optimal control input sequence within multiple future control steps is obtained by solving it. Extract only the control quantity at the first time step. Output execution forms a rolling closed-loop control.

[0132] The optimization problem is described as follows:

[0133] Optimization goal:

[0134] ;

[0135] Constraints:

[0136] ;

[0137] in: These are the state matrix and input matrix of the state-space model; For the first The step disturbance vector can come from external heat flow, material fluctuations, and other influences; To control the set of input constraints, such as upper and lower limits of input power; This is a set of modal state constraints, such as upper and lower temperature limits. This is the optimal control input for the current control cycle, which is then used by the control execution module.

[0138] In general, the optimal control solution unit is solved online using the quadratic programming (QP) algorithm, which has the characteristics of fast solution speed and stable structure.

[0139] In one possible implementation, considering the system with soft constraints (e.g., allowing a certain range of state over-temperature), slack variables can be introduced and the optimization problem can be extended to a penalized soft-constraint QP structure.

[0140] In some further implementations, the control optimization module can be coupled with the sensing feedback module to correct the predicted state based on the actual measured mold temperature information, thereby improving the accuracy and robustness of the prediction model.

[0141] Furthermore, nonlinear factors are often involved in the thermal field control process, and the control optimization module can also be extended to a nonlinear MPC structure. That is, the current state space model is linearized in each control cycle, and nonlinear prediction and optimization algorithms (such as SQP and iLQR) are combined to improve the control accuracy of the high temperature gradient region.

[0142] To address common actuator delay or control dead zone issues in real-world industrial systems, this module also supports dynamic correction through model compensation mechanisms, such as introducing first-order hysteresis models, saturation functions, or dead zone compensation functions, to further align with the actual response characteristics of the system.

[0143] The control execution module is used to drive the mold distributed heating and cooling unit and form a closed-loop control, and to feed back the current mold temperature to the control optimization module.

[0144] In the plastic injection mold control system based on complex curved surfaces and thin-walled structures proposed in this invention, a control execution module is also required to achieve closed-loop drive control of the multi-zone distributed thermal control hardware of the mold and ensure that the control optimization module obtains real-time and effective status feedback information. This module serves as the execution and sensing interface of the entire system. Upstream, it connects to the optimal control input command output by the control optimization module; downstream, it drives the heater and cooler through a hardware interface, simultaneously acquiring the real-time temperature field of the mold and feeding it back to form a closed loop, thus realizing the closed-loop operation and real-time dynamic response of the temperature control system.

[0145] To ensure the integrity of closed-loop control and the effectiveness of the feedback path, the control execution module in this embodiment includes three functional units: a temperature acquisition unit, a state mapping unit, and a signal output unit. The technical solution of the control execution module will be disclosed and supplemented below in conjunction with the foregoing claims, functional logic, and constraint control strategies.

[0146] In this embodiment, the temperature acquisition unit is used to acquire real-time temperature data of different thermal control zones of the mold, providing the original measurement basis for the status feedback of the control optimization module.

[0147] Specifically, the mold is equipped with... There are one temperature sensor, and the measurement point location is recorded as follows: Then at any control moment The collected temperature observation values ​​are:

[0148] ;

[0149] in: This is the measured temperature vector; For a moment mold in position The temperature at that location is expressed in Kelvin (K). To measure the number of channels.

[0150] Generally, the sensor selection needs to meet the requirements of high precision, low drift, and fast response, such as using PT100, thermocouple, or MEMS sensor; the sampling period is set according to the dynamic characteristics of the system, usually in the range of 0.1s to 1s, to ensure the timeliness of the system's dynamic adjustment.

[0151] As an alternative, the sensor layout can be determined by a partitioned symmetrical arrangement or by an optimization algorithm based on thermal response sensitivity, in order to improve the representativeness of the temperature field across the entire mold.

[0152] In order to enable the physical space observations collected by the temperature acquisition unit to correspond to the modal state space defined in the control optimization module, this embodiment sets up a state mapping unit to realize the modal space inverse mapping of the temperature field and output the modal state estimation vector.

[0153] In one specific implementation, the state mapping is constructed using a linear mapping model:

[0154] ;

[0155] in: The estimated modal state vector; This is the state mapping matrix, used to project the sampling space onto the modal space; The aforementioned measured temperature vector; To control the modal dimensions defined in the modeling module.

[0156] In general, matrix It can be constructed in the following two ways:

[0157] Directly construct using the distribution of basis function values ​​at the sensor location (i.e.) ;

[0158] Parameter regression training can be performed using offline data fitting or pseudo-inverse solutions (such as least squares method).

[0159] As an extension, the state mapping unit can also incorporate filtering mechanisms, such as using an extended Kalman filter (EKF) or a sliding window filter, to... Dynamic optimization is performed to further suppress sampling errors or noise interference.

[0160] The signal output unit is used to output the optimal control vector provided by the control optimization module. The signals are parsed and converted into hardware control signals to drive the mold heater and cooling actuator. While ensuring control accuracy, system safety constraints and power limitations must also be met. In this embodiment, the optimal control vector must satisfy the following control amplitude constraints before entering the actuator:

[0161] ;

[0162] Control jump amplitude constraints:

[0163] ;

[0164] in: This is the optimal control vector for the current control step, expressed in W. This is the control vector from the previous control step, in units of W; Minimum and maximum power limits for each actuator, in watts (W); To control the maximum allowable amplitude of the jump, the unit is W; The norm is the infinite norm, representing the absolute value of the largest component of the vector.

[0165] In the signal output unit, if a signal is detected... If any of the above limits are exceeded, a limit-down operation will be performed:

[0166] ;

[0167] And further perform jump limit correction:

[0168] ;

[0169] in: It is a symbolic function; This is the final control command sent to the actuator.

[0170] Control signal types are distinguished according to actuator type:

[0171] For electric heaters: convert to PWM duty cycle signal or analog voltage (0-10V);

[0172] For water-cooled channel valves: the output is either analog current (4-20mA) or digital pulse signal to control the opening degree.

[0173] In practical applications, to enhance execution safety, the signal output unit is also equipped with a status monitoring feedback mechanism to read feedback parameters such as power, current, and valve displacement in real time, so as to judge the effect of command execution and support the future status prediction and update of the control optimization module.

[0174] In some engineering deployments, the control execution module also has actuator group management functions, which can dynamically start and stop some thermal control units according to system settings, such as turning off some heaters when the mold is in a partially idle state to save energy and reduce consumption.

[0175] In addition, this module supports asynchronous parallel communication with the temperature acquisition system, and combined with the clock synchronization strategy, improves the data consistency of the multi-channel thermal control system. When the control loop is interrupted or the system is reset, the control execution module also has fault protection logic, such as temperature limit shutdown and self-test reset, to ensure stable and reliable system operation under actual injection molding conditions.

[0176] Please see Figure 2 The present invention also provides a method for controlling injection molds for plastic parts based on complex curved surfaces and thin-walled structures, comprising the following steps:

[0177] S1. Construct a three-dimensional geometric model of the precision injection mold and divide it into multiple independent or overlapping thermal control regions. Determine the layout of the thermal control units based on the mold's structural characteristics and thermal control requirements. In this step, the spatial mapping relationship between each thermal control region and its corresponding heating / cooling unit is also established, providing a geometric basis for subsequent physical modeling and actuator control.

[0178] S2. Based on the constructed mold geometry, establish a physical model of heat conduction, clarify the thermal properties of materials in each region such as thermal conductivity and heat capacity, set the initial temperature distribution of the mold, and define boundary heat transfer conditions according to the process environment, including convection, radiation, or adiabatic boundaries. This forms a three-dimensional partial differential equation model of heat conduction suitable for unsteady-state conditions.

[0179] S3. Select a set of basis functions that satisfy spatial orthogonality, such as polynomial functions or piecewise linear functions, and use the Galerkin projection method to perform modal dimensionality reduction on the unsteady heat conduction model, transforming it into a finite-dimensional state-space form. This yields the state matrix describing the evolution of the mold's thermal field, the input matrix, and the perturbation term representing external disturbances or heat source distribution.

[0180] S4. Map the target temperature field distribution (such as the expected steady-state temperature gradient or time-varying curve) to a finite-dimensional modal space to obtain the corresponding target modal state vector, which serves as the tracking reference for the controller and is used for state deviation evaluation in the subsequent performance function.

[0181] S5. Construct a performance objective function for control optimization, comprehensively considering factors such as modal state deviation, control input energy consumption, and input change rate. Set corresponding weighting matrices for different control objectives to adjust the contribution of each factor in the optimization process, achieving a balance between control accuracy, energy consumption, and stability.

[0182] S6. Based on the aforementioned state-space model and performance function, the optimal control input sequence for multiple future time steps is solved online using a rolling predictive control strategy (such as model predictive control, MPC), and the control input command required for the current control step is extracted in combination with system constraints.

[0183] S7. By deploying multiple temperature sensors in the key thermal control area of ​​the mold, the current thermal field data is collected in real time, and the sampled temperature vector is converted into the current modal state vector through the state mapping mechanism, which is then used as feedback input to the controller to complete the closed-loop update.

[0184] S8. Based on the optimal control command output by the predictive controller, drive the heating unit and cooling channel actuator in the corresponding area to perform temperature adjustment operation. The system continuously iterates the perception-optimization-execution process to achieve dynamic closed-loop precise control of the thermal field distribution of the precision injection mold.

[0185] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A plastic injection mold control system based on complex curved surfaces and thin-walled structures, characterized in that, include: The mold modeling module is used to construct the three-dimensional geometry of precision injection molds and divide them into multiple independent thermal control areas for the temperature modeling module to establish a heat conduction model. The temperature modeling module is used to establish a heat conduction model of the mold and perform physical modeling, providing thermophysical equation inputs for the control modeling module. The control modeling module is used to perform finite-dimensional modeling transformation of the heat conduction model, providing a state-space model for the control optimization module; The control optimization module is used to construct a predictive controller for the dimensionality-reduced thermal field state system and output control commands to the control execution module. The control execution module is used to drive the mold's distributed heating and cooling unit and form a closed-loop control, and to feed back the current temperature of the mold to the control optimization module.

2. The plastic injection mold control system based on complex curved surfaces and thin-walled structures according to claim 1, characterized in that, The mold modeling module includes: 3D modeling unit, used to construct parametric 3D geometric models of precision injection molds; The region division unit is used to divide the mold into multiple independent thermal control regions based on the mold's functional structure. The control interface calibration unit is used to calibrate the spatial mapping relationship between the thermal control area and the heating and cooling control channel.

3. The plastic injection mold control system based on complex curved surfaces and thin-walled structures according to claim 1, characterized in that, The temperature modeling module includes: The heat conduction modeling unit is used to establish a thermophysical model of precision injection molds based on three-dimensional unsteady heat conduction laws. The parameter setting unit is used to set the thermal conductivity, volumetric heat source distribution, heat transfer boundary conditions, and initial temperature distribution. The boundary condition setting unit is used to establish boundary heat exchange relationships and set boundary heat flow conditions based on the external environment of the mold.

4. The plastic injection mold control system based on complex curved surfaces and thin-walled structures according to claim 1, characterized in that, The control modeling module includes: A basis function construction unit is used to construct a set of spatially orthogonal basis functions to cover the three-dimensional region of the mold. Galerkin projection elements are used to convert a partial differential model of heat conduction into a finite-dimensional ordinary differential state-space model based on the Galerkin method. The state coefficient generation unit is used to generate the state matrix, input matrix, and corresponding modal control variables.

5. The plastic injection mold control system based on complex curved surfaces and thin-walled structures according to claim 4, characterized in that, The finite-dimensional state-space model constructed by the Galerkin projection unit satisfies the following form: ； in: This is the modal state vector; The state matrix; To control the input matrix; To control the input vector; This is a disturbance term.

6. The plastic injection mold control system based on complex curved surfaces and thin-walled structures according to claim 1, characterized in that, The control optimization module includes: The modal target setting unit is used to project the required thermal field distribution information onto a finite-dimensional modal space to obtain the target modal state vector; The performance function construction unit is used to construct a weighted performance objective function that includes temperature deviation, control energy consumption, and control variation terms. The optimal control solution unit is used to solve for the optimal control input sequence based on the prediction model and extract the optimal control command corresponding to the current control step.

7. The plastic injection mold control system based on complex curved surfaces and thin-walled structures according to claim 6, characterized in that, The control performance function constructed by the performance function construction unit Satisfy the following expression: ； in: To predict the length of the time domain; For the first The modal state vector of each prediction step; The target modal state vector corresponds to the ideal thermal field distribution; For the first The control input vector for each control step; This is the weighting matrix for the state error; To control the input weighting matrix; Weighted matrix to control the rate of change.

8. The plastic injection mold control system based on complex curved surfaces and thin-walled structures according to claim 1, characterized in that, The control execution module includes: The temperature acquisition unit is used to acquire real-time temperature data from multiple temperature sensors deployed in the thermal control area of ​​the mold. The state mapping unit is used to convert the acquired temperature data into the corresponding modal space state vector; The signal output unit is used to convert the optimal control command into a control voltage or flow signal and drive the heater or cooling channel actuator in the corresponding area.

9. The plastic injection mold control system based on complex curved surfaces and thin-walled structures according to claim 8, characterized in that, The control input applied by the signal output unit The following constraints must be met: ； ； in: These are the minimum and maximum allowable power for each control unit, respectively; To control the maximum permissible range of change; The infinite norm of a vector is used to limit and control the magnitude of transitions.

10. A method for controlling injection molds for plastic parts based on complex curved surfaces and thin-walled structures, applied to the injection mold control system for plastic parts based on complex curved surfaces and thin-walled structures as described in any one of claims 1-9, characterized in that, Includes the following steps: S1. Construct a three-dimensional geometric model of the precision injection mold and divide it into multiple thermal control zones, and establish the spatial mapping relationship between each zone and the heating and cooling unit; S2. Establish a physical model of heat conduction in precision injection molds, set the thermal conductivity, initial temperature distribution and boundary heat transfer conditions, and form unsteady-state heat conduction equations. S3. Select spatial orthogonal basis functions, perform Galerkin mode dimensionality reduction on the heat conduction equation, construct a finite-dimensional state-space control model, and obtain the modal state matrix, input matrix and disturbance term expression; S4. Map the target temperature field distribution to the modal space and generate the target modal state vector as a control reference; S5. Construct a performance function, which includes a state deviation term, a control input term, and a control change term, and introduce a corresponding weighting matrix; S6. Based on the state-space model and performance function, a predictive control algorithm is used to solve for the optimal control input sequence online, and the control input corresponding to the current control step is extracted. S7. Collect multi-point temperature data of the mold and map it to the modal state space as the current state input controller of the system; S8. Based on the control input obtained from the solution, drive the distributed heating and cooling units to regulate the temperature, thereby realizing dynamic closed-loop control of the thermal field of the precision injection mold.

Images