Intelligent control method and system for powder forming machine

Abstract

The present application relates to the technical field of powder machines, and in particular to an intelligent control method and system for a powder forming machine. The method comprises: obtaining powder material parameters and mold parameters, wherein the powder material parameters include particle size distribution, yield strength, and friction coefficient, and the mold parameters include mold geometry parameters; performing particle dynamics analysis on the powder material parameters based on the discrete element method to generate particle mechanical motion characteristic data, wherein the particle mechanical motion characteristic data includes particle position data, velocity data, contact force data, and force chain distribution data constructed from the contact force data; performing structural mechanical analysis on the mold parameters based on the finite element method to generate mold macroscopic mechanical response data, wherein the mold macroscopic mechanical response data includes stress distribution data, strain distribution data, and displacement distribution data; and generating powder pressing path control instructions for the mold based on the particle mechanical motion characteristic data and the mold macroscopic mechanical response data.

Description

Technical Field

[0001] The present application relates to the technical field of powder machines, and in particular to an intelligent control method and system for a powder forming machine. Background Art

[0002] A powder machine is a mechanical equipment widely used in industrial production. It is mainly used to process various powdered materials, covering multiple production links such as powder mixing, molding, drying, and packaging.

[0003] During the powder molding process, traditional control methods often fail to fully consider the characteristics of the powder material and the mechanical response of the mold during the molding process, resulting in problems such as unstable molding quality and shortened mold life.

[0004] Factors such as the flow and interaction of powder particles, the stress distribution and deformation of the mold have an important influence on the molding effect, but traditional methods are difficult to accurately simulate and control these factors, so they are in urgent need of improvement. Summary of the Invention

[0005] Based on this, it is necessary to provide an intelligent control method and system for a powder forming machine that can improve the pressing effect of the powder machine in order to address the above technical problems.

[0006] In a first aspect, the present application provides an intelligent control method for a powder forming machine, the method comprising:

[0007] Obtaining powder material parameters and mold parameters, wherein the powder material parameters include particle size distribution, yield strength, and friction coefficient, and the mold parameters include mold geometry parameters;

[0008] Based on the discrete element method, the particle dynamics analysis of the powder material parameters is performed to generate the particle mechanical motion characteristic data, which includes the particle position data, velocity data, contact force data and the force chain distribution data constructed by the contact force data;

[0009] Perform structural mechanics analysis on mold parameters based on the finite element method to generate mold macroscopic mechanical response data, which includes stress distribution data, strain distribution data, and displacement distribution data;

[0010] According to the mechanical motion characteristic data of the particles and the macroscopic mechanical response data of the mold, the powder pressing path control instructions of the mold are generated.

[0011] In one embodiment, a particle dynamics analysis is performed on powder material parameters based on a discrete element method to generate particle mechanical motion characteristic data, including:

[0012] Powder particles are modeled as a collection of discrete particles with particle size distribution and friction coefficient;

[0013] Establish a particle-to-particle contact mechanics model that includes normal contact force and tangential friction force, and define the contact boundary conditions between the particles and the mold wall;

[0014] The composite force on each particle is iteratively calculated within the discrete time step of the pressing process. The composite force includes gravity, contact force between particles and force on the mold wall.

[0015] According to the updated position data and speed data of the particles, until the pressing process is terminated;

[0016] The particle motion trajectory data is extracted and a force chain network is constructed. The force chain network includes the spatial direction, strength and topological structure characteristic parameters of the force chain.

[0017] In one embodiment, a structural mechanics analysis of mold parameters is performed based on the finite element method to generate macroscopic mechanical response data of the mold, including:

[0018] Construct a three-dimensional parametric model based on the mold geometry parameters and perform finite element meshing;

[0019] Define the elastic-plastic constitutive model of the mold material. The model parameters include elastic modulus, Poisson's ratio and dynamic yield strength.

[0020] The powder-mold interaction force field data is converted into a distributed load on the mold surface and applied to the corresponding boundary of the finite element model;

[0021] Solve the finite element control equations including material nonlinearity to obtain the stress data, strain data and displacement distribution data of the mold.

[0022] In one embodiment, the method further comprises:

[0023] Generate stress cloud map, strain cloud map and displacement vector map of key sections of the mold based on the mold's stress data, strain data and displacement distribution data;

[0024] Calculate the maximum equivalent stress value, maximum plastic strain value and maximum displacement;

[0025] When the maximum equivalent stress value is less than the dynamic yield strength of the mold material, the stress data, strain data, and displacement distribution data of the mold are determined to have passed the verification.

[0026] In one embodiment, generating a powder pressing path control instruction for a mold based on the particle mechanical motion characteristic data and the mold macroscopic mechanical response data includes:

[0027] The contact force data in the particle mechanical motion characteristic data and the stress distribution data in the mold macroscopic mechanical response data are coupled and calculated to determine the powder-mold interaction force field data;

[0028] Based on the collaborative analysis of particle mechanical motion characteristic data, mold macroscopic mechanical response data and powder-mold interaction force field data, the powder pressing path control instructions of the mold are generated.

[0029] In one embodiment, coupling calculation is performed on contact force data in the particle mechanical motion characteristic data and stress distribution data in the mold macroscopic mechanical response data to determine the powder-mold interaction force field data, including:

[0030] Establish the mapping relationship between discrete element particle contact force and finite element node force, and use Gaussian interpolation method to convert discrete contact force into continuous distributed load

[0031] Through bidirectional coupling iterative calculation, the contact force distribution of the discrete element system and the stress distribution of the finite element system meet the force balance condition;

[0032] Record the spatial distribution data of the interaction force field when the convergence condition is reached.

[0033] In one embodiment, based on the collaborative analysis of particle mechanical motion characteristic data, mold macroscopic mechanical response data, and powder-mold interaction force field data, powder pressing path control instructions for the mold are generated, including:

[0034] Construct a multi-dimensional feature matrix including force chain network features, mold deformation features and interaction force field features;

[0035] A multi-objective optimization of the pressing path parameters was performed based on a genetic algorithm to generate an optimized pressing path parameter set including the pressing speed curve, pressure loading curve, and holding time. The multiple optimization objectives included minimizing the maximum equivalent stress of the mold, maximizing the uniformity of the powder compaction density, and minimizing the pressing energy consumption.

[0036] According to the optimized pressing path parameter set, the powder pressing path control instructions of the mold are generated.

[0037] In one embodiment, generating powder pressing path control instructions for a mold according to an optimized pressing path parameter set includes:

[0038] Acquire real-time sensing data of the mold, including mold surface pressure distribution, punch displacement and mold temperature;

[0039] Based on real-time sensor data, calculate the deviation dynamics between the current suppression state and the target state;

[0040] According to the deviation dynamics, the pressing path parameter set is adjusted to obtain the optimal pressure path;

[0041] Generate powder pressing path control instructions for the mold based on the optimal pressure path.

[0042] In one embodiment, adjusting the pressing path parameter set according to the deviation dynamics to obtain the optimal pressure path includes:

[0043] The model predictive control algorithm is used to adjust the path parameters of the pressing path parameter set according to the deviation dynamics to obtain the optimal pressure path; the path parameters include pressure gradient, pressurization timing and holding time.

[0044] In a second aspect, the present application also provides an intelligent control system for a powder forming machine, comprising:

[0045] an acquisition module, for acquiring powder material parameters and mold parameters, wherein the powder material parameters include particle size distribution, yield strength and friction coefficient, and the mold parameters include mold geometric parameters;

[0046] The powder analysis module is used to perform particle dynamics analysis on powder material parameters based on the discrete element method to generate particle mechanical motion characteristic data. The particle mechanical motion characteristic data includes particle position data, velocity data, contact force data, and force chain distribution data constructed from the contact force data.

[0047] The mold analysis module is used to perform structural mechanics analysis on mold parameters based on the finite element method and generate mold macro-mechanical response data, which includes stress distribution data, strain distribution data, and displacement distribution data;

[0048] The path instruction generation module is used to generate the powder pressing path control instructions of the mold based on the particle mechanical motion characteristic data and the mold macro-mechanical response data.

[0049] The intelligent control method and system for the powder forming machine described above comprehensively considers the particle mechanical motion characteristic data and the mold macromechanical response data to generate powder pressing path control instructions, enabling precise control of the powder forming process. Based on the powder flow and compaction requirements, parameters such as pressing speed, pressure, and holding time are rationally adjusted to ensure uniform filling and compaction of the powder in the mold, thereby improving the density uniformity, dimensional accuracy, and mechanical properties of the molded product, reducing the occurrence of molding defects, and improving the stability of molding quality.

[0050] By combining powder material properties, mold mechanical response, and pressing path control, the system achieves intelligent control of the powder forming process. Based on real-time data, the system automatically adjusts pressing parameters to accommodate varying powder material and mold conditions, improving production efficiency and product quality stability while reducing errors caused by manual intervention and empirical judgment. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following briefly introduces the drawings required for use in the embodiments of the present application or related technical descriptions. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying any creative work.

[0052] Figure 1 Schematic diagram of a flow chart of an intelligent control method for a powder forming machine in one embodiment;

[0053] Figure 2 A schematic flow chart of the steps for generating particle mechanical motion characteristic data in one embodiment;

[0054] Figure 3 A schematic flow chart of the steps for generating macroscopic mechanical response data of a mold in one embodiment;

[0055] Figure 4 FIG1 is a flow chart of the steps of determining the force field data of the powder-mold interaction in one embodiment;

[0056] Figure 5 A schematic flow chart of the steps of generating powder pressing path control instructions for a mold in one embodiment;

[0057] Figure 6 The figure is a flow chart of the steps of generating powder pressing path control instructions for a mold in one embodiment. DETAILED DESCRIPTION

[0058] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0059] In an exemplary embodiment, Figure 1 As shown, an intelligent control method for a powder forming machine is provided, comprising:

[0060] S101, obtaining powder material parameters and mold parameters.

[0061] Among them, the powder material parameters include particle size distribution, yield strength and friction coefficient, and the mold parameters include mold geometry parameters.

[0062] Alternatively, the particle size distribution can be measured using a laser particle size analyzer, where the distribution follows the Rosin-Rammler distribution. This distribution can well describe the statistical characteristics of the powder particle size and provide a basis for subsequent discrete element analysis.

[0063] Optionally, the yield strength and friction coefficient can be calibrated using a material testing machine. The yield strength reflects the stress at which the powder material begins to plastically deform, while the friction coefficient affects the relative motion between powder particles and between the powder and the die.

[0064] Optionally, mold parameters are mainly mold geometric parameters, such as cavity size, wall thickness, etc. These parameters are obtained through the 3D CAD model, which can accurately describe the shape and structure of the mold and provide an accurate geometric model for finite element analysis.

[0065] S102, performing particle dynamics analysis on the powder material parameters based on the discrete element method to generate particle mechanical motion characteristic data.

[0066] The particle mechanical motion characteristic data include particle position data, velocity data, contact force data and force chain distribution data constructed by the contact force data.

[0067] The discrete element method (DEM) is a numerical method used to simulate the dynamic behavior of particle systems. It treats powder as consisting of a large number of discrete particles, each of which obeys Newton's laws of motion. By considering factors such as contact forces between particles and gravity, it simulates the movement and interaction of particles during the molding process.

[0068] In the particle mechanical motion data, position data records the spatial position of each particle at different times, reflecting the particle's motion trajectory during the molding process. Velocity data indicates the velocity of each particle at different times, including its magnitude and direction, which helps analyze the particle's motion state. Contact force data describes the magnitude and direction of contact forces between particles and between particles and the mold, and is key data for analyzing particle interactions and force transmission. Force chain distribution data is constructed from contact force data and reflects the force transmission path and network structure between particles.

[0069] S103, performing structural mechanics analysis on the mold parameters based on the finite element method to generate macroscopic mechanical response data of the mold.

[0070] The macroscopic mechanical response data of the mold include stress distribution data, strain distribution data and displacement distribution data.

[0071] It can be understood that the finite element method is a numerical analysis method that discretizes a continuum into a finite number of units. By analyzing and calculating the mechanical properties of each unit, the mechanical response of the entire structure is obtained.

[0072] Within the mold's macroscopic mechanical response data, stress distribution data indicates the magnitude and direction of stress experienced at different locations within the mold, helping to determine whether the mold will experience strength failure. Strain distribution data reflects the degree of deformation of the mold under stress. Analyzing strain distribution helps optimize mold design and improve molding accuracy. Displacement distribution data describes the displacement of various locations within the mold and can be used to assess the mold's overall deformation and stability.

[0073] S104 , generating a powder pressing path control instruction of the mold according to the particle mechanical motion characteristic data and the mold macroscopic mechanical response data.

[0074] Optionally, preprocess the particle mechanical motion characteristic data and mold macromechanical response data, including data cleaning, filtering, and normalization, to improve data quality and usability. Key features are then extracted from this data, such as the powder's bulk density, fluidity index, the mold's maximum stress point, and the maximum deformation area. These features can more concisely describe the mechanical behavior of the powder and mold, providing a basis for subsequent analysis and decision-making. For example:

[0075] Through theoretical analysis, numerical simulation, or experimental research, a mapping relationship or mathematical model can be established between the mechanical motion characteristics of the particles, the macroscopic mechanical response of the mold, and the powder compaction path. For example, a prediction model can be constructed based on machine learning algorithms such as neural networks and decision trees. This model inputs data on the mechanical motion characteristics of the particles and the macroscopic mechanical response of the mold and outputs the optimal powder compaction path parameters. This model can learn the complex nonlinear relationships between the data, thereby achieving accurate predictions and decisions.

[0076] Alternatively, when generating control instructions for the powder compaction path, multiple optimization objectives must be considered, such as maximizing powder compaction density, minimizing die stress, and improving molding efficiency. These objectives may conflict with each other, necessitating the use of multi-objective optimization algorithms, such as genetic algorithms and particle swarm optimization, to find an optimal set of compaction path parameters that balances all objectives while satisfying certain constraints. For example, while ensuring that the powder compaction density meets the required level, the maximum die stress can be minimized to extend the die's service life.

[0077] Based on the optimized pressing path parameters, specific powder pressing path control instructions are generated. These instructions usually include the settings of parameters such as pressing speed, pressure, pressing time, and holding time.

[0078] In an exemplary embodiment, Figure 2 As shown in the figure, the particle dynamics analysis of powder material parameters is performed based on the discrete element method to generate particle mechanical motion characteristic data, including:

[0079] S201, modeling the powder particles as a discrete particle set with a particle size distribution and a friction coefficient.

[0080] The powder particles are modeled as a collection of discrete particles with a specific particle size distribution and friction coefficient. The particle size distribution is determined using previously acquired measurement data and typically follows a Rosin-Rammler distribution. The friction coefficient is obtained by calibrating a materials testing machine. This modeling approach more realistically reflects the actual state of the powder particles.

[0081] S202, establishing an inter-particle contact mechanics model including normal contact force and tangential friction force, and defining contact boundary conditions between the particles and the mold wall.

[0082] Optionally, a particle-to-particle contact mechanics model can be constructed that includes normal contact force and tangential friction. The normal contact force describes the force generated when particles squeeze each other, while the tangential friction accounts for the resistance between particles when they slide relative to each other.

[0083] Common contact mechanics models include the Hertz-Mindlin model, which can accurately simulate the contact behavior between particles.

[0084] Furthermore, the contact boundary conditions between the particles and the mold wall are defined to clarify the mechanical behavior of the particles when in contact with the mold wall, such as the friction coefficient and elastic recovery coefficient between the particles and the mold wall, to ensure that the simulation results are consistent with the actual situation.

[0085] S203, iteratively calculating the composite force of each particle within the discrete time step of the pressing process.

[0086] The composite forces include gravity, contact force between particles and force acting on the mold wall.

[0087] Optionally, the composite force on each particle is iteratively calculated within a discrete time step of the pressing process. The composite force on the particle includes gravity, contact force between particles, and force acting on the mold wall.

[0088] It's understandable that gravity is an inherent property of the particles themselves. The contact force between particles is calculated based on the established contact mechanics model, while the force acting on the mold wall is determined based on the contact between the particles and the mold wall and the boundary conditions. Through continuous iterative calculations, the force applied to the particles can be updated in real time.

[0089] S204, updating the position data and speed data of the particles until the pressing process is terminated.

[0090] Optionally, the position data and velocity data of the particle are updated according to Newton's law of motion (F=ma, where F is the net force acting on the particle, m is the mass of the particle, and a is the acceleration of the particle).

[0091] Optionally, within each discrete time step, the acceleration is calculated based on the force acting on the particle, thereby obtaining the change in velocity and position. This process is repeated until the pressing process is completed. In this way, the trajectory of the particle throughout the pressing process can be simulated.

[0092] S205, extracting particle motion trajectory data and constructing a force chain network, where the force chain network includes characteristic parameters of the force chain's spatial direction, strength, and topological structure.

[0093] Optionally, the motion trajectory data of the particles are extracted from the above calculation process. These data record the position information of the particles at different moments and reflect the motion path of the particles.

[0094] A force chain network is constructed based on the contact force data between particles. The force chain network is a network structure composed of particles that contact and transmit force to each other, and contains characteristic parameters such as the spatial direction, strength, and topological structure of the force chain.

[0095] It can be understood that the spatial direction of the force chain describes the direction of force transmission between particles, the force chain strength indicates the magnitude of the force, and the topological structural characteristic parameters reflect the overall structure and connection mode of the force chain network. By analyzing the force chain network, we can gain a deeper understanding of the force transmission mechanism and mechanical properties within the powder particles.

[0096] In this embodiment, through precise modeling and a reasonable contact mechanics model, the dynamic behavior of powder particles during the pressing process can be simulated more accurately, providing a reliable basis for the optimization of the powder forming process.

[0097] In an exemplary embodiment, Figure 3 As shown, the structural mechanics analysis of the mold parameters is performed based on the finite element method to generate the mold macroscopic mechanical response data, including:

[0098] S301, constructing a three-dimensional parametric model according to the mold geometric parameters and performing finite element meshing.

[0099] Optionally, based on the acquired mold geometry, a 3D parametric model of the mold can be constructed using specialized finite element analysis software. This 3D parametric model accurately reflects the mold's actual shape and dimensions. This model is then meshed using finite element methods, discretizing the continuous mold structure into a finite number of elements and nodes for numerical calculations. The quality and density of the meshing affect the accuracy and efficiency of the calculation results and should be appropriately configured based on the mold's complexity and analysis requirements.

[0100] S302, defining an elastic-plastic constitutive model of the mold material, where the model parameters include elastic modulus, Poisson's ratio, and dynamic yield strength.

[0101] Optionally, define an elastic-plastic constitutive model of the mold material, which can describe the elastic and plastic deformation behaviors of the mold material under load.

[0102] Model parameters include elastic modulus, Poisson's ratio, and dynamic yield strength. The elastic modulus reflects a material's ability to resist deformation within its elastic range; the Poisson's ratio describes the relationship between lateral and longitudinal strains when a material is subjected to stress; and the dynamic yield strength indicates the stress at which plastic deformation begins. These parameters can be obtained through material testing or by referring to relevant literature.

[0103] S303 , converting the powder-mold interaction force field data into a mold surface distributed load, and applying it to the corresponding boundary of the finite element model.

[0104] Optionally, the die interaction force field data can be converted into distributed die surface loads. The powder-die interaction force field data reflects the distribution of forces exerted by the powder on the die surface during compaction. Using a specific conversion method, these forces are converted into surface distributed loads suitable for finite element analysis and applied to the corresponding boundaries of the finite element model. This simulates the forces acting on the die during actual powder compaction.

[0105] S304, solving the finite element control equation including material nonlinearity to obtain the stress data, strain data and displacement distribution data of the mold.

[0106] Optionally, solve the finite element governing equations that include material nonlinearity. Because mold materials may undergo plastic deformation when subjected to stress, resulting in nonlinear stress-strain relationships, the finite element governing equations that account for material nonlinearity must be solved.

[0107] By solving these equations, the stress data, strain data and displacement distribution data of the mold can be obtained.

[0108] Furthermore, the method also includes: generating stress cloud maps, strain cloud maps and displacement vector maps of key sections of the mold based on the stress data, strain data and displacement distribution data of the mold; calculating the maximum equivalent stress value, maximum plastic strain value and maximum displacement; when the maximum equivalent stress value is less than the dynamic yield strength of the mold material, determining that the stress data, strain data and displacement distribution data of the mold have passed the verification.

[0109] Among them, according to the stress data, strain data and displacement distribution data of the mold, the stress cloud map, strain cloud map and displacement vector map of the key section of the mold are generated.

[0110] Stress cloud maps visually display the magnitude and distribution of stress in different parts of the mold, with darker colors indicating greater stress. Strain cloud maps reflect the degree of mold deformation, and displacement vector maps show the direction and magnitude of displacement at each point in the mold. These graphs help engineers quickly understand the mold's mechanical response and identify potential problem areas.

[0111] Among them, the maximum equivalent stress value reflects the maximum stress level that the mold is subjected to; the maximum plastic strain value indicates the maximum degree of plastic deformation of the mold; and the maximum displacement reflects the maximum deformation of the mold.

[0112] When the maximum equivalent stress value is less than the dynamic yield strength of the mold material, the mold's stress, strain, and displacement distribution data are considered verified. This means that under the current stress conditions, the mold will not undergo plastic deformation and meets the design requirements. If the maximum equivalent stress value exceeds the dynamic yield strength of the mold material, the mold design needs to be optimized, such as adjusting the mold geometry or changing the material, and then re-analyzed and verified.

[0113] In this embodiment, through accurate finite element analysis and data verification, potential problems of the mold can be discovered in the design stage, and optimization can be carried out in a timely manner to reduce the failure rate of the mold in actual use.

[0114] In an exemplary embodiment, generating a powder pressing path control instruction of a mold based on the particle mechanical motion characteristic data and the mold macroscopic mechanical response data includes:

[0115] The contact force data in the particle mechanical motion characteristic data and the stress distribution data in the mold macromechanical response data are coupled and calculated to determine the powder-mold interaction force field data; based on the collaborative analysis of the particle mechanical motion characteristic data, the mold macromechanical response data and the powder-mold interaction force field data, the powder pressing path control instructions of the mold are generated.

[0116] Optionally, contact force data can be selected from the particle mechanical motion characteristics data. This data reflects the magnitude and direction of the interaction forces between powder particles and between particles and the mold. Simultaneously, stress distribution data can be selected from the mold macroscopic mechanical response data. This data reflects the stress experienced by various parts of the mold during the powder compaction process.

[0117] The contact force data is coupled with the stress distribution data. This is because the contact forces of the powder particles directly act on the mold, affecting the mold's stress distribution; the mold's stress state, in turn, influences the movement and contact of the powder particles. By comprehensively considering the mutual influence of the two, the coupled calculation can more accurately determine the powder-mold interaction force field data. This force field data comprehensively describes the interaction between the powder and the mold throughout the entire pressing process, including information such as the magnitude, direction, and distribution of the force.

[0118] Furthermore, a collaborative analysis was conducted based on the particle mechanical motion data, the mold macromechanical response data, and the powder-mold interaction force field data. The particle mechanical motion data allows us to understand the motion trajectory and velocity changes of the powder particles during the pressing process; the mold macromechanical response data helps us understand the stress, strain, and displacement distribution of the mold; and the powder-mold interaction force field data reveals the interaction mechanism between the two.

[0119] Based on the results of the collaborative analysis, control instructions for the powder compaction path are generated for the mold. These instructions precisely define various parameters of the powder forming machine during the compaction process, such as compaction speed, pressure, direction, and duration. For example, if the analysis reveals excessive stress in a particular area of ​​the mold, the instructions may adjust the compaction speed or pressure to prevent excessive force in that area. If powder particles are unevenly distributed in certain areas, the instructions may change the compaction direction to promote even powder filling.

[0120] In this embodiment, by precisely considering the movement of powder particles and the mechanical response of the mold, the generated pressing path control instructions can enable the powder to fill the mold cavity more evenly, reduce the density differences and defects of the molded product, and thus improve the quality and performance of the product.

[0121] In an exemplary embodiment, Figure 4 As shown, the contact force data in the particle mechanical motion characteristic data and the stress distribution data in the mold macroscopic mechanical response data are coupled and calculated to determine the powder-mold interaction force field data, including:

[0122] S401, establish a mapping relationship between discrete element particle contact force and finite element node force, and use Gaussian interpolation method to convert discrete contact force into continuous distributed load.

[0123] Alternatively, in the discrete element method (DEM), the contact forces between powder particles are discretely distributed, whereas the finite element method (FEM) requires continuously distributed loads for analyzing mold mechanical response. Therefore, the mapping between the DEM particle contact forces and the FEM node forces must first be established. This step acts like a bridge, transferring the discrete particle contact force information to the nodes of the FEM model.

[0124] Gaussian interpolation is used to convert discrete contact forces into continuously distributed loads. Gaussian interpolation is a commonly used numerical interpolation method that calculates the function value at any point within a continuous region by weighted averaging the function value at discrete points. Here, Gaussian interpolation is used to convert discrete particle contact forces into continuously distributed loads, which are then applied to the finite element model. This allows the finite element model to accurately simulate the mechanical response of the mold under the action of powder contact forces.

[0125] S402, through bidirectional coupling iterative calculation, the contact force distribution of the discrete element system and the stress distribution of the finite element system meet the force balance condition.

[0126] Alternatively, the interaction between powder and mold is a dynamic process. The movement and contact forces of powder particles affect the stress distribution of the mold, while the stress distribution of the mold in turn affects the movement and contact state of powder particles. Therefore, a bidirectional coupled iterative calculation is required.

[0127] During the iterative calculation process, the contact force distribution of the discrete element system and the stress distribution of the finite element system are ensured to satisfy the force equilibrium condition. The force equilibrium condition is a fundamental principle in physics, stating that when an object is in equilibrium under the action of a force, the net force is zero. In the case of powder-mold interaction, this requires ensuring that the sum of the contact forces of the powder particles in the discrete element system and the sum of the stresses experienced by the mold in the finite element system are balanced in all directions. Through continuous iterative adjustments, the two are gradually brought closer to a state of force equilibrium.

[0128] S403, recording the spatial distribution data of the interaction force field when the convergence condition is reached.

[0129] Optionally, you need to set a convergence condition during the bidirectionally coupled iterative calculation. This condition typically means that the iterative calculation is considered to have converged when the difference between the contact force distribution of the discrete element system and the stress distribution of the finite element system is less than a preset threshold. The setting of this threshold depends on the specific application scenario and the required computational accuracy.

[0130] When convergence conditions are reached, the spatial distribution data of the interaction force field is recorded. This data, including the magnitude, direction, and spatial distribution of the interaction force between the powder and the die, is key to describing the powder-die interaction force field and provides an important basis for the subsequent generation of powder compaction path control instructions.

[0131] In this embodiment, by establishing a mapping relationship and using bidirectional coupled iterative calculation, the discrete characteristics of the powder particles and the continuous structural characteristics of the mold are fully considered, so that the calculation results more accurately reflect the actual situation of the powder-mold interaction force field.

[0132] In an exemplary embodiment, Figure 5 As shown, based on the collaborative analysis of particle mechanical motion characteristic data, mold macroscopic mechanical response data and powder-mold interaction force field data, the powder pressing path control instructions of the mold are generated, including:

[0133] S501, constructing a multi-dimensional feature matrix including force chain network features, mold deformation features and interaction force field features.

[0134] Optionally, relevant features of the force chain network, such as the length distribution, strength distribution, and topological structure of the force chain, can be extracted from the discrete element analysis results. Mold deformation data can be obtained from the finite element analysis results to calculate mold deformation characteristics such as maximum deformation and deformation distribution. Interaction force field characteristics, such as force magnitude, direction, and distribution, can be extracted from the powder-mold interaction force field data. These features are arranged according to specific rules into a multidimensional feature matrix, where each row or column represents a feature vector.

[0135] S502 , performing multi-objective optimization on the pressing path parameters based on a genetic algorithm to generate an optimized pressing path parameter set including a pressing speed curve, a pressure loading curve, and a holding time.

[0136] Among them, the multiple optimization objectives include minimizing the maximum equivalent stress of the mold, maximizing the uniformity of powder compaction density and minimizing the compaction energy consumption.

[0137] Optionally, define the value range of the pressing path parameters, including the value range of the pressing speed curve, the value range of the pressure loading curve, and the value range of the holding time. Initialize a set of random pressing path parameters as the initial population of the genetic algorithm. For each individual (i.e., a set of pressing path parameters), simulation calculations are performed based on the multi-dimensional feature matrix to evaluate its performance in the three goals of minimizing the maximum equivalent stress of the mold, maximizing the uniformity of the powder pressing density, and minimizing the pressing energy consumption, and calculate the fitness value. Use the selection, crossover, and mutation operations of the genetic algorithm to iteratively update the population and continuously search for a better combination of pressing path parameters. When the preset number of iterations or convergence conditions are met, stop the iteration and select the individual with the best fitness value as the optimized pressing path parameter set.

[0138] S503: Generate powder pressing path control instructions for the mold according to the optimized pressing path parameter set.

[0139] Optionally, based on the optimized compaction path parameter set, parameters such as the compaction speed curve, pressure loading curve, and dwell time are converted into specific control instruction codes. These instruction codes can be transmitted to the control system of the powder forming machine via an interface to control the powder forming machine to perform powder compaction operations according to the optimized path.

[0140] In an exemplary embodiment, Figure 6 As shown, according to the optimized pressing path parameter set, the powder pressing path control instructions of the mold are generated, including:

[0141] S601, obtaining real-time sensing data of the mold.

[0142] Among them, real-time sensing data includes mold surface pressure distribution, punch displacement and mold temperature.

[0143] Optionally, a pressure sensor is installed on the mold surface to measure the pressure distribution on the mold surface; a displacement sensor is installed on the punch to monitor the punch displacement; and a temperature sensor is installed at the key position of the mold to obtain the mold temperature in real time.

[0144] S602, based on the real-time sensor data, calculating the deviation dynamics between the current suppression state and the target state.

[0145] Optionally, based on the optimized pressing path parameter set, the target state of the powder pressing process is determined, such as the target pressure distribution, the target punch displacement, and the target die temperature.

[0146] The real-time sensor data acquired from the mold is compared with the target state, and the deviation dynamics between the current pressing state and the target state are calculated. The deviation dynamics reflects the degree of deviation and change trend between the actual pressing process and the expected target.

[0147] S603: Adjust the pressure path parameter set according to the deviation dynamics to obtain the optimal pressure path.

[0148] Optionally, a model predictive control algorithm is used, which can optimize and adjust control parameters based on the current state of the system and future prediction information. Based on the calculated deviation dynamics, the model predictive control algorithm is used to adjust the path parameters of the suppression path parameter set.

[0149] Path parameters include pressure gradient, pressurization sequence, and hold time. By adjusting these parameters, the pressing process can quickly approach the target state and obtain the optimal pressure path. For example, if the pressure distribution on the mold surface is uneven, the pressure gradient can be adjusted to achieve a more uniform pressure distribution. If the punch displacement deviates significantly from the target value, the pressurization sequence can be adjusted to ensure that the punch moves according to the expected displacement.

[0150] S604: Generate powder pressing path control instructions for the mold according to the optimal pressure path.

[0151] Optionally, the optimal pressure path is converted into a specific control instruction code, and the instruction is sent to the control system of the powder molding machine through the communication interface to control the powder molding machine to perform powder pressing operations according to the optimal path.

[0152] Specifically, according to the deviation dynamics, the pressing path parameter set is adjusted to obtain the optimal pressure path, including: using a model predictive control algorithm to adjust the path parameters of the pressing path parameter set according to the deviation dynamics to obtain the optimal pressure path; the path parameters include pressure gradient, pressurization timing and holding time.

[0153] As you can understand, Model Predictive Control (MPC) is an advanced model-based control strategy. It predicts the future behavior of a system and calculates optimal control inputs based on the predicted results and set objectives. In the powder compaction scenario, the algorithm uses the system's dynamic model, combined with current deviation dynamic information, to predict the state changes of the compaction process over a period of time. It then optimizes and calculates the control parameters that will bring the system closer to the target state.

[0154] Before using a model predictive control algorithm, a dynamic model of the powder compaction process must be established. This model describes the relationship between compaction path parameters (pressure gradient, pressurization sequence, and dwell time) and mold conditions (die surface pressure distribution, punch displacement, and die temperature). This model can be established through theoretical analysis, experimental data fitting, or machine learning. For example, regression analysis can be used to establish a mathematical relationship between pressure gradient and die surface pressure distribution using extensive experimental data.

[0155] The calculated deviation dynamics between the current pressing state and the target state serve as input to the model predictive control algorithm. This deviation dynamics includes information such as the current magnitude of the deviation and its trend over time. For example, if the current deviation in the mold surface pressure distribution is large and increasing, this indicates that the pressing process is deviating increasingly from the target state, necessitating more aggressive algorithm adjustments.

[0156] Based on the established system model and current deviation dynamics, the algorithm predicts the state of the pressing process over a period of time. The prediction horizon is a key parameter, determining the timeframe over which the algorithm predicts future states. A longer prediction horizon can consider longer-term impacts but increases computational complexity; a shorter prediction horizon focuses more on near-term state changes. For example, it can predict changes in die surface pressure distribution, punch displacement, and die temperature over the next several pressing cycles.

[0157] Based on the predicted future state and the set optimization goal, the algorithm adjusts the path parameters of the suppression path parameter set through optimization calculation. The specific adjusted path parameters include:

[0158] The pressure gradient determines the rate at which pressure increases or decreases during the compaction process. If the current pressure distribution on the mold surface is uneven, with some areas of low pressure, the algorithm may increase the pressure gradient, causing the pressure to rise faster to improve compaction in those areas. Conversely, if some areas of high pressure are too high, the algorithm may decrease the pressure gradient.

[0159] The pressurization sequence specifies the timing and order in which pressure is applied at different stages. If the punch displacement deviates significantly from the target value, the algorithm may adjust the pressurization sequence, advancing or delaying the pressurization operation to ensure that the punch moves according to the expected displacement.

[0160] The dwell time is the duration that pressure is maintained after reaching the set pressure. If the mold temperature is too high or the powder compaction effect is not ideal, the algorithm may extend the dwell time to increase the compaction density of the powder. If the compaction effect is already good, the dwell time may be shortened to improve production efficiency.

[0161] The model predictive control algorithm is an iterative process. Within each control cycle, predictions and optimization calculations are re-performed based on the latest deviation dynamics, continuously adjusting path parameters to ensure the pressing process consistently approaches the target state. This real-time dynamic adjustment effectively addresses various uncertainties in the production process, ensuring the stability of the powder pressing process and the quality of the final product.

[0162] Through continuous adjustments to the model predictive control algorithm, the optimal pressure path is ultimately determined, allowing the pressing process to approach the target state. This optimal pressure path takes into account both current deviation dynamics and future predictions, providing accurate operational guidance for the powder forming machine.

[0163] It should be understood that, although the steps in the flowcharts of the above embodiments are shown in sequence as indicated by the arrows, these steps are not necessarily performed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be performed in other orders. Moreover, at least a portion of the steps in the flowcharts of the above embodiments may include multiple steps or multiple stages, and these steps or stages are not necessarily performed at the same time, but can be performed at different times. The execution order of these steps or stages is not necessarily to be performed in sequence, but can be performed in turn or alternately with other steps or at least a portion of steps or stages in other steps.

[0164] Based on the same inventive concept, embodiments of the present application also provide an intelligent control system for a powder forming machine for implementing the aforementioned intelligent control method for a powder forming machine. The solution provided by this device is similar to the solution described in the aforementioned method. Therefore, the specific limitations of one or more embodiments of the intelligent control system for a powder forming machine provided below can be found in the aforementioned limitations of the intelligent control method for a powder forming machine, and will not be further elaborated here.

[0165] In an exemplary embodiment, an intelligent control system for a powder forming machine includes:

[0166] an acquisition module, for acquiring powder material parameters and mold parameters, wherein the powder material parameters include particle size distribution, yield strength and friction coefficient, and the mold parameters include mold geometric parameters;

[0167] The powder analysis module is used to perform particle dynamics analysis on powder material parameters based on the discrete element method to generate particle mechanical motion characteristic data. The particle mechanical motion characteristic data includes particle position data, velocity data, contact force data, and force chain distribution data constructed from the contact force data.

[0168] The mold analysis module is used to perform structural mechanics analysis on mold parameters based on the finite element method and generate mold macro-mechanical response data, which includes stress distribution data, strain distribution data, and displacement distribution data;

[0169] The path instruction generation module is used to generate the powder pressing path control instructions of the mold based on the particle mechanical motion characteristic data and the mold macro-mechanical response data.

[0170] Each module in the intelligent control system of the powder forming machine described above may be implemented in whole or in part through software, hardware, or a combination thereof. Each module may be embedded in or independent of a processor in a computer device in the form of hardware, or may be stored in a computer device memory in the form of software, so that the processor can call and execute the corresponding operations of each module.

[0171] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. In particular, any reference to memory, database, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The databases involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the various embodiments provided herein may be, but are not limited to, general-purpose processors, central processing units (CPUs), graphics processing units (GPUs), digital signal processors (DSPs), programmable logic devices (PLDs), quantum computing-based data processing logic devices, artificial intelligence (AI) processors, and the like.

[0172] The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0173] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, and these modifications and improvements fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. An intelligent control method for a powder forming machine, characterized in that: The method comprises: Acquiring powder material parameters and mold parameters, wherein the powder material parameters include particle size distribution, yield strength, and friction coefficient, and the mold parameters include mold geometry parameters; Performing particle dynamics analysis on the powder material parameters based on the discrete element method to generate particle mechanical motion characteristic data, wherein the particle mechanical motion characteristic data includes particle position data, velocity data, contact force data, and force chain distribution data constructed from the contact force data; Performing structural mechanics analysis on the mold parameters based on the finite element method to generate mold macroscopic mechanical response data, wherein the mold macroscopic mechanical response data includes stress distribution data, strain distribution data, and displacement distribution data; generating a powder pressing path control instruction of the mold according to the particle mechanical motion characteristic data and the mold macroscopic mechanical response data; The step of generating a powder pressing path control instruction for a mold according to the particle mechanical motion characteristic data and the mold macroscopic mechanical response data includes: Couple the contact force data in the particle mechanical motion characteristic data with the stress distribution data in the mold macroscopic mechanical response data to determine the powder-mold interaction force field data; Generate powder pressing path control instructions for the mold based on collaborative analysis of the particle mechanical motion characteristic data, the mold macroscopic mechanical response data, and the powder-mold interaction force field data; The coupling calculation of the contact force data in the particle mechanical motion characteristic data and the stress distribution data in the mold macroscopic mechanical response data to determine the powder-mold interaction force field data includes: The mapping relationship between discrete element particle contact force and finite element node force is established, and the discrete contact force is converted into a continuous distributed load using Gaussian interpolation method; Through bidirectional coupling iterative calculation, the contact force distribution of the discrete element system and the stress distribution of the finite element system meet the force balance condition; Record the spatial distribution data of the interaction force field when the convergence condition is reached; The generating of the powder pressing path control instructions of the mold based on the collaborative analysis of the particle mechanical motion characteristic data, the mold macroscopic mechanical response data and the powder-mold interaction force field data includes: Construct a multi-dimensional feature matrix including force chain network features, mold deformation features and interaction force field features; A multi-objective optimization of the pressing path parameters was performed based on a genetic algorithm to generate an optimized pressing path parameter set including the pressing speed curve, pressure loading curve, and holding time. The multiple optimization objectives included minimizing the maximum equivalent stress of the mold, maximizing the uniformity of the powder compaction density, and minimizing the pressing energy consumption. generating a powder pressing path control instruction for the mold according to the optimized pressing path parameter set; The step of generating a powder pressing path control instruction for a mold according to the optimized pressing path parameter set includes: Acquiring real-time sensing data of the mold, wherein the real-time sensing data includes mold surface pressure distribution, punch displacement, and mold temperature; Calculating the deviation dynamics between the current suppression state and the target state based on the real-time sensor data; According to the deviation dynamics, adjusting the pressing path parameter set to obtain an optimal pressure path; According to the optimal pressure path, a powder pressing path control instruction of the mold is generated.

2. The method according to claim 1, characterized in that The particle dynamics analysis of the powder material parameters based on the discrete element method to generate particle mechanical motion characteristic data includes: Modeling the powder particles as a collection of discrete particles having the particle size distribution and friction coefficient; Establish a particle-to-particle contact mechanics model that includes normal contact force and tangential friction force, and define the contact boundary conditions between the particles and the mold wall; Iteratively calculate the composite force on each particle within the discrete time step of the pressing process; the composite force includes gravity, contact force between particles and force acting on the mold wall; According to the updated position data and speed data of the particles, until the pressing process is terminated; The particle motion trajectory data is extracted and a force chain network is constructed, wherein the force chain network includes characteristic parameters of the force chain space direction, strength and topological structure.

3. The method according to claim 1, characterized in that The structural mechanics analysis of the mold parameters based on the finite element method to generate mold macroscopic mechanical response data includes: Constructing a three-dimensional parametric model according to the mold geometric parameters and performing finite element meshing; Define the elastic-plastic constitutive model of the mold material. The model parameters include elastic modulus, Poisson's ratio and dynamic yield strength. Converting the powder-mold interaction force field data into a mold surface distributed load and applying it to the corresponding boundary of the finite element model; Solve the finite element control equations including material nonlinearity to obtain the stress data, strain data and displacement distribution data of the mold.

4. The method according to claim 3, characterized in that The method further comprises: Generate a stress nephogram, a strain nephogram, and a displacement vector diagram of a key section of the mold based on the stress data, strain data, and displacement distribution data of the mold; Calculate the maximum equivalent stress value, maximum plastic strain value and maximum displacement; When the maximum equivalent stress value is less than the dynamic yield strength of the mold material, it is determined that the stress data, strain data, and displacement distribution data of the mold have passed the verification.

5. The method according to claim 1, characterized in that The step of adjusting the pressing path parameter set according to the deviation dynamics to obtain the optimal pressure path includes: A model predictive control algorithm is used to adjust the path parameters of the pressing path parameter set according to the deviation dynamics to obtain the optimal pressure path; the path parameters include pressure gradient, pressurization timing and holding time.

6. An intelligent control system for a powder forming machine, characterized in that: The system comprises: an acquisition module, configured to acquire powder material parameters and mold parameters, wherein the powder material parameters include particle size distribution, yield strength, and friction coefficient, and the mold parameters include mold geometry parameters; a powder analysis module for performing particle dynamics analysis on the powder material parameters based on a discrete element method to generate particle mechanical motion characteristic data, wherein the particle mechanical motion characteristic data includes particle position data, velocity data, contact force data, and force chain distribution data constructed from the contact force data; A mold analysis module, configured to perform structural mechanics analysis on the mold parameters based on a finite element method to generate mold macroscopic mechanical response data, wherein the mold macroscopic mechanical response data includes stress distribution data, strain distribution data, and displacement distribution data; a path instruction generating module, configured to generate a powder pressing path control instruction for the mold according to the particle mechanical motion characteristic data and the mold macroscopic mechanical response data; The step of generating a powder pressing path control instruction for a mold according to the particle mechanical motion characteristic data and the mold macroscopic mechanical response data includes: Couple the contact force data in the particle mechanical motion characteristic data with the stress distribution data in the mold macroscopic mechanical response data to determine the powder-mold interaction force field data; Generate powder pressing path control instructions for the mold based on collaborative analysis of the particle mechanical motion characteristic data, the mold macroscopic mechanical response data, and the powder-mold interaction force field data; The coupling calculation of the contact force data in the particle mechanical motion characteristic data and the stress distribution data in the mold macroscopic mechanical response data to determine the powder-mold interaction force field data includes: The mapping relationship between discrete element particle contact force and finite element node force is established, and the discrete contact force is converted into a continuous distributed load using Gaussian interpolation method; Through bidirectional coupling iterative calculation, the contact force distribution of the discrete element system and the stress distribution of the finite element system meet the force balance condition; Record the spatial distribution data of the interaction force field when the convergence condition is reached; The generating of the powder pressing path control instructions of the mold based on the collaborative analysis of the particle mechanical motion characteristic data, the mold macroscopic mechanical response data and the powder-mold interaction force field data includes: Construct a multi-dimensional feature matrix including force chain network features, mold deformation features and interaction force field features; A multi-objective optimization of the pressing path parameters was performed based on a genetic algorithm to generate an optimized pressing path parameter set including the pressing speed curve, pressure loading curve, and holding time. The multiple optimization objectives included minimizing the maximum equivalent stress of the mold, maximizing the uniformity of the powder compaction density, and minimizing the pressing energy consumption. generating a powder pressing path control instruction for the mold according to the optimized pressing path parameter set; The step of generating a powder pressing path control instruction for a mold according to the optimized pressing path parameter set includes: Acquiring real-time sensing data of the mold, wherein the real-time sensing data includes mold surface pressure distribution, punch displacement, and mold temperature; Calculating the deviation dynamics between the current suppression state and the target state based on the real-time sensor data; According to the deviation dynamics, adjusting the pressing path parameter set to obtain an optimal pressure path; According to the optimal pressure path, a powder pressing path control instruction of the mold is generated.

Images