A precision CNC injection molding system

Abstract

The present invention relates to the field of CNC injection molding technology, and in particular to a precision CNC injection molding system. It includes an injection molding data acquisition module, a precision injection molding module, an injection molding process control module, a defect monitoring module and a feedback optimization module. The injection molding data acquisition module obtains key data in the injection molding process through sensors. The precision injection molding module generates a three-dimensional model based on these data, and uses finite element analysis to perform simulation optimization to generate optimal injection molding parameters. The injection molding process control module adjusts key process parameters in real time through a differential dynamic programming algorithm. The defect monitoring module monitors and identifies defect types in real time, and the feedback optimization module performs adaptive optimization based on defect analysis to form a closed-loop control to continuously improve the injection molding quality. The present invention not only improves the precision and injection molding quality of granular building blocks, optimizes the matching precision between granular building blocks, but also reduces the defect rate and improves production efficiency.

Description

Technical Field

[0001] The invention relates to the technical field of numerical control injection molding, in particular to a precision numerical control injection molding system. Background Art

[0002] As a common children's toy, plastic particle building blocks are widely used to develop children's intelligence, exercise hands-on ability and cultivate spatial imagination. This type of building blocks is usually made of polymer materials through injection molding, requiring smooth appearance, precise size, firm splicing and certain pressure resistance. However, traditional injection molding technology often faces some technical bottlenecks when processing plastic particle building blocks, such as insufficient production accuracy, fast mold wear, low material utilization rate and poor product consistency. At present, there are still the following problems: the existing technology lacks efficient geometric modeling and simulation optimization methods when performing plastic particle building block injection molding, and usually relies on empirical parameters for debugging, which makes it difficult to achieve accurate three-dimensional modeling and process optimization, resulting in difficulty in obtaining the best process parameters during the injection molding process, and the final product quality is unstable; traditional injection molding process control mostly adopts fixed parameter settings, lacks intelligent adjustment capabilities, and cannot be dynamically adjusted according to the real-time changing injection molding situation, resulting in the inability to accurately control key parameters, thereby affecting the stability of the injection molding process and the qualified rate of the finished product. Summary of the invention

[0003] To solve the above problems, the present invention provides a precision CNC injection molding system, which solves the problem of how to achieve accurate three-dimensional modeling and process optimization in the injection molding process of plastic particle building blocks, and dynamically adjust key process parameters to improve product quality and production stability, thereby improving the precision and injection molding quality of the particle building blocks and optimizing the matching accuracy between the particle building blocks.

[0004] To achieve the above object, the technical solution adopted by the present invention is:

[0005] A precision CNC injection molding system, comprising an injection molding data acquisition module, a precision injection molding module, an injection molding process control module, a defect monitoring module and a feedback optimization module connected in sequence;

[0006] The injection molding data acquisition module is used to collect the injection molding data of the plastic particle building blocks through sensors;

[0007] The precision injection molding module is used to generate a three-dimensional model of the plastic particle building block based on the injection molding data using a geometric modeling algorithm, and to simulate and optimize the injection molding process in combination with finite element analysis and numerical simulation algorithms, and to generate optimal injection molding parameters;

[0008] The injection molding process control module is used to control and adjust the key process parameters in the injection molding process in real time based on the optimal injection molding parameters using a differential dynamic programming algorithm, and dynamically plan the optimal injection molding process control path; the key process parameters include injection molding pressure, temperature, flow rate and time;

[0009] The defect monitoring module is used to monitor the defects of plastic particle building blocks in the injection molding process in real time and determine the defect types, use variational mode decomposition algorithm and machine learning algorithm to identify and extract potential defects in the injection molding process, and generate a defect analysis report; the defect types include bubbles, shrinkage marks, warpage, cracks and material delamination;

[0010] The feedback optimization module is used to perform adaptive feedback optimization based on the optimal injection molding process control path and the defect analysis report using a reinforcement learning algorithm to form a closed-loop control and continuously optimize the injection molding quality.

[0011] Furthermore, the injection molding data includes the size, melting point, viscosity, injection molding pressure, injection molding temperature, injection molding flow rate and injection molding time of the plastic particle building blocks.

[0012] Furthermore, the operation process of the precision injection molding module includes the following steps:

[0013] Preprocessing the injection molding data, and generating a three-dimensional model of the plastic particle building block using a Bezier curve adaptive subdivision algorithm;

[0014] Using finite element analysis algorithms, the thermal-mechanical coupling effect, stress distribution and material flow of the model during the injection molding process are dynamically simulated;

[0015] Combined with the results of finite element analysis, numerical simulation algorithms are used to dynamically simulate and optimize the material fluidity, curing time, temperature changes and other process parameters during the injection molding process;

[0016] Based on the finite element analysis and numerical simulation results, the optimal injection molding parameters are generated.

[0017] Furthermore, the construction process of the finite element analysis algorithm includes the following steps:

[0018] Based on the three-dimensional model of plastic particle building blocks, the model is meshed using adaptive meshing technology;

[0019] According to the physical and mechanical properties of the plastic particles, corresponding material parameters are set; the material parameters include density, elastic modulus, thermal conductivity, viscosity and specific heat capacity of the plastic particles;

[0020] Combined with the injection molding process requirements of plastic particle building blocks, reasonable boundary conditions and initial conditions are set. The boundary conditions include injection pressure, mold temperature and flow channel design, and the initial conditions include the setting of initial temperature field and velocity field.

[0021] According to the thermal-mechanical coupling effect and nonlinear behavior of materials involved in the injection molding process, the implicit finite element method is used, and the multi-step iteration technology is combined to solve the complex injection molding process;

[0022] After solving the simulation, the calculation results are analyzed using post-processing tools to generate data on the temperature distribution, stress-strain distribution, and flow path of the plastic particles during the injection molding process.

[0023] Furthermore, the operation process of the injection molding process control module includes the following steps:

[0024] Based on the optimal injection molding parameters and preliminary key process parameters generated by the precision injection molding module, an initial injection molding process path is formed;

[0025] Through the differential dynamic programming algorithm, based on real-time monitoring data, dynamic adjustment and optimization are performed on key process parameters to generate a real-time updated injection molding process control path;

[0026] Use iterative optimization algorithms to dynamically plan and optimize the injection molding path based on historical injection molding data and current process status;

[0027] The optimized process path is continuously executed during the injection molding process, and the process parameters are adjusted through the feedback optimization module in combination with real-time defect monitoring data and analysis results to form a closed-loop control system.

[0028] Furthermore, the formula of the differential dynamic programming algorithm is as follows:

[0029]

[0030] in, From time t to The minimum cost function of the optimal injection molding process control path is used to optimize the process parameters in the injection molding process; Indicates the key process parameters in the injection molding process; Indicates the system status at different process stages during the injection molding process; Represents the loss or deviation in the injection molding process at each time t; Indicates the time point when injection molding is completed; Indicates the system status at the end of injection molding.

[0031] Furthermore, the operation process of the defect monitoring module includes the following steps:

[0032] Acquiring injection molding data from the injection molding data acquisition module;

[0033] Based on the variational modal decomposition algorithm, the injection molding data is subjected to time-frequency decomposition, the complex injection molding process signal is decomposed into multiple modes with different frequency and time characteristics, and the key process characteristic signals reflecting the injection molding process are extracted;

[0034] The characteristic signals extracted by variational modal decomposition are input into the pulse coupled neural network. Combined with the spatial and temporal coupling characteristics of the network, the characteristic signals of different modes are multi-dimensionally fused. By adaptively adjusting the neuron connection weights, the potential defect types in the injection molding process can be accurately identified and a defect analysis report can be generated.

[0035] Furthermore, the formula of the variational mode decomposition algorithm is as follows:

[0036]

[0037] in, Represents different frequency features extracted from the injection molding process signal; represents the kth modal signal The corresponding phase function; represents the original input signal, i.e., the signal sequence that changes over time; It represents the components with specific frequency and time domain characteristics separated by the variational mode decomposition algorithm; Represents the modal signal The corresponding phase function; K represents the total number of modes; represents the adjustment parameter; Represents the modal phase gradient.

[0038] Furthermore, the feedback optimization module realizes real-time update of optimization algorithm model parameters of large-scale injection molding process data based on the cloud database, and automatically iterates and optimizes the control solution according to the big data analysis results in the production line.

[0039] The beneficial effects of the present invention are:

[0040] The injection molding data acquisition module of the present invention collects key data in the injection molding process in real time through sensors, ensuring the reliability and accuracy of the precise modeling and subsequent optimization of the plastic particle building blocks. The precision injection molding module uses algorithms such as geometric modeling and finite element analysis to accurately model the plastic particle building blocks in three dimensions, and combines numerical simulation algorithms for simulation optimization, thereby generating optimal injection molding parameters and improving the accuracy and stability of the injection molding process. The injection molding process control module uses a differential dynamic programming algorithm to dynamically adjust the key parameters in the injection molding process in real time, realizes intelligent control of the injection molding process, significantly improves the stability and consistency of the injection molding process, and reduces errors caused by human intervention. The defect monitoring module uses variational mode decomposition and machine learning algorithms to monitor defects in the injection molding process in real time and determine the defect type. By generating a defect analysis report, it provides a detailed basis for subsequent optimization, thereby improving product quality and reducing scrap rate. The feedback optimization module combines the defect analysis report and the optimal injection molding process control path, and uses the reinforcement learning algorithm to perform adaptive feedback optimization to form a closed-loop control. It continuously self-adjusts and optimizes the process parameters during the injection molding process, realizes self-learning and continuous improvement of the injection molding system, improves the precision and injection quality of the granular building blocks, and optimizes the matching accuracy between the granular building blocks. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a module schematic diagram of a precision CNC injection molding system of the present invention.

[0042] Figure 2 It is a flowchart of a finite element analysis algorithm construction process provided by an embodiment of the present invention.

[0043] Figure 3 It is a flow chart of the operation process of the injection molding process control module provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0044] See also Figure 1-3 As shown, the present invention relates to a precision numerically controlled injection molding system.

[0045] Example

[0046] A precision CNC injection molding system, comprising an injection molding data acquisition module, a precision injection molding module, an injection molding process control module, a defect monitoring module and a feedback optimization module connected in sequence;

[0047] The injection molding data acquisition module is used to collect the injection molding data of the plastic particle building blocks through sensors; the injection molding data includes the size, melting point, viscosity, injection molding pressure, injection molding temperature, injection molding flow rate and injection molding time of the plastic particle building blocks;

[0048] It should be noted that the sensor is specifically as follows:

[0049] Temperature sensor: Temperature sensors are used to monitor key temperature data of plastic pellets from heating to injection molding. Specifically, the temperature sensor will be installed inside the screw barrel of the injection molding machine, and multiple temperature sensors will be distributed from the feed port to the nozzle to monitor the heating process of the molten plastic. These sensors will ensure that the temperature gradient of the heating section matches the melting point of the material to prevent overheating or insufficient heating from causing unstable melt quality. Temperature sensors are arranged in different areas of the mold to monitor the mold temperature during the cooling process. By sensing the cooling speed of the mold, it is ensured that the building blocks will not deform or have excessive internal stress when solidifying.

[0050] Pressure sensor: During the injection molding process, pressure fluctuations will directly affect the molding quality of the final product. Pressure sensors are mainly installed when the screw pushes the plastic melt. The pressure sensor can record the changes in injection force in real time to ensure that the plastic maintains a uniform injection pressure when entering the mold. Before the plastic melt enters the mold, the pressure sensor can record the pressure value at the entrance to detect the smoothness of the melt flow and whether there are abnormal pressure fluctuations, which helps to avoid bubbles or other structural defects in the product.

[0051] Viscosity sensor: Viscosity is an important parameter that affects the fluidity of plastic melt, which directly determines whether the material can fill the mold smoothly. The online viscosity sensor detects the rheological properties of the melt in real time, especially for plastic raw materials with multiple additives or different particle sizes. Installed at the end of the screw, it evaluates the change in viscosity by controlling and monitoring the shear rate of the melt. The data will be fed back to the control system to adjust the injection speed and pressure in real time to ensure that the viscosity is within the ideal range.

[0052] Flow rate sensor: Flow rate sensor is used to monitor the flow rate of plastic melt during the injection molding process. Specifically, the flow rate sensor will be deployed at the injection nozzle outlet to measure the flow rate of the plastic melt when it leaves the injection nozzle, ensuring that the melt has a uniform flow rate when entering the mold, and will not cause uneven flow or premature solidification due to too fast or too slow flow rate.

[0053] Dimension monitoring system: Through non-contact measurement tools (such as laser rangefinders or optical sensors), the size of plastic granule building blocks is monitored in real time to ensure that the finished product is consistent with the designed size. The system will be installed at the end of the production line to detect the shape, holes and structure of the building blocks to detect dimensional deviations in time.

[0054] In actual operation, the data acquisition module ensures the accuracy and real-time nature of the data through a multi-level processing flow, and the injection molding data acquisition module integrates and processes the data from different sensors. For example, the real-time data of temperature, pressure and flow rate sensors are summarized through a unified data processing platform. Based on built-in rules and algorithms, the platform filters out outliers and noise to ensure that the collected data has high accuracy. Due to the different sampling frequencies of different sensors, the system synchronizes various types of data through timestamps. In particular, for parameter changes in high-dynamic processes, such as melt flow rate and pressure, data at the same time point need to be strictly aligned to ensure the accuracy of subsequent simulation analysis and optimization control. The data acquisition module stores the collected real-time data in the local storage to ensure that the system can still keep data records when the network is delayed or interrupted. At the same time, the module uploads the data to the cloud database through a high-speed transmission protocol (such as MQTT or HTTP protocol) for subsequent optimization and machine learning algorithm training. Cloud data can also be compared and analyzed with historical production data to evaluate whether the current production status meets historical process standards.

[0055] The data acquisition module provides a user-friendly visual interface, where operators can view the changes of various data in real time through a large screen or mobile device. The interface displays real-time values ​​of temperature, pressure, flow rate and viscosity in the form of charts and dashboards, and marks the safety range and deviations of the current operation. When certain key parameters (such as pressure or temperature) exceed the preset range, the system will issue an alarm to prompt the operator to intervene. The alarm information will be notified through the screen, sound or mobile device to ensure timely response. The system allows operators to trace back and analyze historical injection molding data to help identify potential problems. For example, by comparing multiple production data, operators can analyze why a batch of products has defects and find the reasons for related process parameter fluctuations.

[0056] The injection molding data acquisition module can be connected with other intelligent devices in the entire production line through the Internet of Things to form a highly automated production environment. Through IoT technology, the system can seamlessly connect with the control module, quality monitoring equipment and feedback optimization system of the injection molding machine to form a closed-loop control process. The integrated 5G communication module and industrial Ethernet interface ensure high data throughput and low latency data transmission. This is especially important for large-scale real-time data analysis and remote control.

[0057] The precision injection molding module is used to generate a three-dimensional model of the plastic particle building block based on the injection molding data using a geometric modeling algorithm, and to simulate and optimize the injection molding process in combination with finite element analysis and numerical simulation algorithms, and to generate optimal injection molding parameters;

[0058] The operation process of the precision injection molding module includes the following steps:

[0059] Preprocessing the injection molding data, and generating a three-dimensional model of the plastic particle building block using a Bezier curve adaptive subdivision algorithm;

[0060] Specifically, the geometric modeling step includes reading the basic geometric data of the particle building blocks, including structural information such as size, shape, holes and protrusions. Based on this data, the Bezier curve algorithm subdivides the curve part, especially near the edges and holes of the building blocks. The subdivided patches can capture the geometric features more accurately. A 3D model with an adaptive mesh is generated, which maintains geometric accuracy and improves computational efficiency.

[0061] Using finite element analysis algorithms, the thermal-mechanical coupling effect, stress distribution and material flow of the model during the injection molding process are dynamically simulated;

[0062] Combined with the results of finite element analysis, numerical simulation algorithms are used to dynamically simulate and optimize the material fluidity, curing time, temperature changes and other process parameters during the injection molding process;

[0063] Generate optimal injection molding parameters based on finite element analysis and numerical simulation results;

[0064] The construction process of the finite element analysis algorithm includes the following steps:

[0065] Based on the three-dimensional model of plastic particle building blocks, the model is meshed using adaptive meshing technology;

[0066] According to the physical and mechanical properties of the plastic particles, corresponding material parameters are set; the material parameters include density, elastic modulus, thermal conductivity, viscosity and specific heat capacity of the plastic particles;

[0067] Combined with the injection molding process requirements of plastic particle building blocks, reasonable boundary conditions and initial conditions are set. The boundary conditions include injection pressure, mold temperature and flow channel design, and the initial conditions include the setting of initial temperature field and velocity field.

[0068] According to the thermal-mechanical coupling effect and nonlinear behavior of materials involved in the injection molding process, the implicit finite element method is used, and the multi-step iteration technology is combined to solve the complex injection molding process;

[0069] Specifically, the implicit finite element method is used to analyze the thermomechanical coupling effect to ensure accurate simulation of the interaction between temperature, stress and strain during the injection molding process. The system simulates the entire process of material heating, injection molding and cooling in stages, especially the shrinkage and stress release behavior of the material during cooling. During the injection molding process, the plastic melt exhibits nonlinear behavior, especially the rheological properties and stress-strain relationship at high temperatures. Through multi-step iterative technology, the system can accurately simulate these complex nonlinear characteristics, thereby better predicting the flow path, curing time and stress distribution of the material.

[0070] After solving the simulation, the calculation results are analyzed using post-processing tools to generate data on the temperature distribution, stress-strain distribution, and flow path of the plastic particles during the injection molding process.

[0071] The formula of the three-dimensional model is as follows:

[0072] in, Represents the three-dimensional coordinate point of the plastic particle building block model corresponding to the parameters (u, v); The control vertex array representing the plastic particle building block is used to define the geometric shape of the plastic particle building block model; It represents the deformation function, reflecting the influence of dynamic changes of temperature, pressure and material flow rate on geometric deformation during the injection molding process; and Represents the basis functions of a Bezier surface.

[0073] The injection molding process control module is used to control and adjust the key process parameters in the injection molding process in real time based on the optimal injection molding parameters using a differential dynamic programming algorithm, and dynamically plan the optimal injection molding process control path; the key process parameters include injection molding pressure, temperature, flow rate and time;

[0074] The operation process of the injection molding process control module includes the following steps:

[0075] Based on the optimal injection molding parameters and preliminary key process parameters generated by the precision injection molding module, an initial injection molding process path is formed;

[0076] Through the differential dynamic programming algorithm, based on real-time monitoring data, dynamic adjustment and optimization are performed on key process parameters to generate a real-time updated injection molding process control path;

[0077] Specifically, the input of the differential dynamic programming algorithm includes the real-time process data of the injection molding machine, such as real-time injection pressure, temperature, flow rate, etc., and the material state data fed back by the sensor (such as melt temperature, mold cooling temperature, etc.). The entire injection molding process is divided into multiple state nodes, such as filling, holding pressure, cooling and other stages. The state of each node is described by the process parameters and system feedback data at that time. In the actual production process, the system monitors the key process parameters in real time through sensors and inputs these real-time data into the differential dynamic programming algorithm. The process path is dynamically adjusted according to the current state (such as small fluctuations in pressure or temperature). For example, when the system detects that the melt flow rate slows down due to temperature changes during the filling process, the differential dynamic programming algorithm will adjust the injection pressure and flow rate in real time to maintain the stability of the filling and prevent the occurrence of flow dead zones. The system takes reducing defects (such as bubbles, shrinkage marks, etc.) and improving production efficiency as optimization goals. The differential dynamic programming algorithm optimizes the objective function by gradient descent to find the optimal process control path while ensuring product quality.

[0078] Use iterative optimization algorithms to dynamically plan and optimize the injection molding path based on historical injection molding data and current process status;

[0079] The optimized process path is continuously executed during the injection molding process, and the process parameters are adjusted through the feedback optimization module in combination with real-time defect monitoring data and analysis results to form a closed-loop control system.

[0080] Furthermore, the formula of the differential dynamic programming algorithm is as follows:

[0081]

[0082] in, From time t to The minimum cost function of the optimal injection molding process control path is used to optimize the process parameters in the injection molding process; Indicates the key process parameters in the injection molding process; Indicates the system status at different process stages during the injection molding process; Represents the loss or deviation in the injection molding process at each time t; Indicates the time point when injection molding is completed; Indicates the system status at the end of injection molding.

[0083] The defect monitoring module is used to monitor the defects of plastic particle building blocks in the injection molding process in real time and determine the defect types, use variational mode decomposition algorithm and machine learning algorithm to identify and extract potential defects in the injection molding process, and generate a defect analysis report; the defect types include bubbles, shrinkage marks, warpage, cracks and material delamination;

[0084] The operation process of the defect monitoring module includes the following steps:

[0085] Acquiring injection molding data from the injection molding data acquisition module;

[0086] Based on the variational modal decomposition algorithm, the injection molding data is subjected to time-frequency decomposition, the complex injection molding process signal is decomposed into multiple modes with different frequency and time characteristics, and the key process characteristic signals reflecting the injection molding process are extracted;

[0087] Specifically, the specific steps of the variational mode decomposition algorithm (VMD) are as follows: The system processes the original signal during the injection molding process and decomposes the complex signal into multiple separate modes. For example, the pressure signal may contain different frequency components, representing the characteristics of different stages of the injection molding process (such as rapid filling, holding pressure and cooling). VMD divides these signals into multiple modes according to frequency, and each mode represents a different process feature. After each modal signal is decomposed, the system extracts time and frequency features, which reflect the changes in key process parameters during the injection molding process. For example, the frequency feature can be used to identify whether the injection pressure is stable, and the time feature can be used to determine whether the temperature change in the cooling stage is uniform. The system extracts characteristic signals closely related to defects from the decomposed modes, such as: Low-frequency signal: represents the overall pressure change during the injection molding process. If there is a large fluctuation, it may cause warping or shrinkage marks. High-frequency signal: reflects the rapid filling state of the material in the mold. If the filling is uneven, it may cause bubbles or material stratification.

[0088] The characteristic signals extracted by variational modal decomposition are input into the pulse coupled neural network. Combined with the spatial and temporal coupling characteristics of the network, the characteristic signals of different modes are multi-dimensionally fused. By adaptively adjusting the neuron connection weights, the potential defect types in the injection molding process can be accurately identified. According to the severity and impact of the defects, the defects are classified and prioritized, and a defect analysis report is generated.

[0089] Furthermore, the formula of the variational mode decomposition algorithm is as follows:

[0090]

[0091] in, Represents different frequency features extracted from the injection molding process signal, that is, different components of the signal in different time periods, which are used to capture the specific frequency or time-varying characteristics of the signal; represents the kth modal signal The corresponding phase function describes the phase change of the modal signal on the time axis, reflects the phase characteristics of the signal, and can help identify defects or anomalies in the injection molding process; represents the original input signal, i.e., the signal sequence that changes over time; It represents the components with specific frequency and time domain characteristics separated by the variational mode decomposition algorithm; Represents the modal signal The corresponding phase function describes the phase of the modal signal changing with time; K represents the total number of modes; represents a tuning parameter used to control the sparsity of modal decomposition. It affects the smoothness of each modal signal. The larger the value, the smoother the generated modal signal; Represents the modal phase The gradient of the phase, the rate of change of the phase with time.

[0092] The feedback optimization module is used to perform adaptive feedback optimization based on the optimal injection molding process control path and defect analysis report using a reinforcement learning algorithm to form a closed-loop control and continuously optimize the injection molding quality; the feedback optimization module realizes real-time update optimization algorithm model parameters of large-scale injection molding process data based on a cloud database, and automatically iterates the optimization control solution based on the big data analysis results in the production line.

[0093] It should be noted that the feedback optimization module realizes a fully automatic control process from data collection to process optimization through a closed-loop control system. When the defect monitoring module detects a defect in the product (such as bubbles, cracks, etc.), the feedback optimization module will immediately start the optimization process. First, the system will find historical data related to the defect, analyze the process parameters used when similar defects occurred in the past, and generate a preliminary adjustment plan. The feedback optimization module adjusts the process parameters in the next injection molding process in real time based on the defect monitoring results and historical data analysis. For example, if the system detects shrinkage marks on the product, it may be caused by too short cooling time or insufficient pressure. The system will automatically increase the cooling time or increase the pressure to avoid the same defect in the next product. The system combines the real-time data collected by the sensor with the optimization model in the cloud to form a closed-loop control process. The data in each production is used as the input for the next production cycle, and the process parameters are continuously optimized through the reinforcement learning algorithm to ensure that the injection molding process is always in the optimal state. The closed-loop control system can automatically adjust the parameters without human intervention, greatly improving production efficiency and product quality. The feedback optimization module has a self-healing function. If the system finds that a certain parameter always deviates from the optimal range during multiple production runs, it will prioritize optimizing this parameter by adjusting the algorithm weight. For example, when the system detects product warping multiple times, it will automatically prioritize optimizing parameters such as temperature and pressure, and trace back historical data to find the most likely cause, thereby repairing the process defect.

[0094] In summary, the present invention uses a variety of sensors (including temperature, pressure, viscosity, flow rate, etc.) to enable the system to accurately and in real time collect the injection molding data of plastic particle building blocks, ensuring the stability of key process parameters during the production process. The multi-point arrangement of sensors and data synchronization technology further ensure high-precision data acquisition, providing a reliable basis for subsequent process optimization. Using the precision injection molding module, based on finite element analysis and numerical simulation, the system can simulate and optimize the injection molding process in advance, generate optimal injection molding parameters, and effectively control the fluidity, curing time and stress distribution of the material, and reduce the probability of defects such as bubbles, warping, and shrinkage marks.

[0095] The injection molding process control module in the present invention uses a differential dynamic programming algorithm to monitor and adjust key process parameters (such as injection pressure, temperature, flow rate, etc.) in real time, dynamically plan the optimal injection molding path, and ensure the stability of the production process and the consistency of the product. At the same time, the feedback optimization module combines real-time defect monitoring data to achieve closed-loop control, further improving the stability of production and the quality of finished products. The defect monitoring module uses variational mode decomposition and machine learning algorithms to identify potential defect types in the injection molding process in real time, analyze and generate defect reports. Through accurate defect classification and priority sorting, the system can adjust production parameters in a timely manner, reduce rework or waste caused by defects, and improve production efficiency.

[0096] The present invention integrates the Internet of Things technology and can be linked with other intelligent devices to achieve highly automated production line management. At the same time, users can monitor various process data in real time through a friendly visual interface and respond promptly when an abnormality occurs, which improves the convenience of operation and the transparency of the production process.

[0097] The above implementation modes are merely descriptions of the preferred implementation modes of the present invention, and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary engineering and technical personnel in the field shall fall within the protection scope determined by the claims of the present invention.

Claims

1. A precision CNC injection molding system, characterized in that: It includes an injection molding data acquisition module, a precision injection molding module, an injection molding process control module, a defect monitoring module and a feedback optimization module which are connected in sequence; The injection molding data acquisition module is used to collect the injection molding data of the plastic particle building blocks through sensors; The precision injection molding module is used to generate a three-dimensional model of the plastic particle building block based on the injection molding data using a geometric modeling algorithm, and to simulate and optimize the injection molding process in combination with finite element analysis and numerical simulation algorithms, and to generate optimal injection molding parameters; The injection molding process control module is used to control and adjust the key process parameters in the injection molding process in real time based on the optimal injection molding parameters using a differential dynamic programming algorithm, and dynamically plan the optimal injection molding process control path; the key process parameters include injection molding pressure, temperature, flow rate and time; The defect monitoring module is used to monitor the defects of plastic particle building blocks in the injection molding process in real time and determine the defect types, use variational mode decomposition algorithm and machine learning algorithm to identify and extract potential defects in the injection molding process, and generate a defect analysis report; the defect types include bubbles, shrinkage marks, warpage, cracks and material delamination; The feedback optimization module is used to perform adaptive feedback optimization based on the optimal injection molding process control path and the defect analysis report using a reinforcement learning algorithm to form a closed-loop control and continuously optimize the injection molding quality; The operation process of the injection molding process control module includes the following steps: Based on the optimal injection molding parameters and preliminary key process parameters generated by the precision injection molding module, an initial injection molding process path is formed; Through the differential dynamic programming algorithm, based on real-time monitoring data, dynamic adjustment and optimization are performed on key process parameters to generate a real-time updated injection molding process control path; Use iterative optimization algorithms to dynamically plan and optimize the injection molding path based on historical injection molding data and current process status; Continuously execute the optimized process path during the injection molding process, and adjust the process parameters through the feedback optimization module, combined with real-time defect monitoring data and analysis results, to form a closed-loop control system; The formula of the differential dynamic programming algorithm is as follows: ； in, From time t to The minimum cost function of the optimal injection molding process control path is used to optimize the process parameters in the injection molding process; Indicates the key process parameters in the injection molding process; Indicates the system status at different process stages during the injection molding process; Represents the loss or deviation in the injection molding process at each time t; Indicates the time point when injection molding is completed; Indicates the system status at the end of injection molding; The operation process of the defect monitoring module includes the following steps: Acquiring injection molding data from the injection molding data acquisition module; Performing time-frequency decomposition on the injection molding data based on a variational modal decomposition algorithm, decomposing a complex injection molding process signal into multiple modes with different frequency and time characteristics, and extracting key process characteristic signals reflecting the injection molding process; The characteristic signals extracted by variational mode decomposition are input into the pulse coupled neural network. The characteristic signals of different modes are multi-dimensionally fused by combining the spatial and temporal coupling characteristics of the network. By adaptively adjusting the neuron connection weights, the potential defect types in the injection molding process can be accurately identified and a defect analysis report can be generated. The formula of the variational mode decomposition algorithm is as follows: ； in, Represents different frequency features extracted from the injection molding process signal; represents the kth modal signal The corresponding phase function; represents the original input signal, i.e., the signal sequence that changes over time; It represents the components with specific frequency and time domain characteristics separated by the variational mode decomposition algorithm; Represents the modal signal The corresponding phase function; K represents the total number of modes; represents the adjustment parameter; Represents the modal phase gradient.

2. A precision CNC injection molding system according to claim 1, characterized in that: The injection molding data includes the size, melting point, viscosity, injection molding pressure, injection molding temperature, injection molding flow rate and injection molding time of the plastic particle building blocks.

3. A precision CNC injection molding system according to claim 1, characterized in that: The operation process of the precision injection molding module includes the following steps: Preprocessing the injection molding data, and generating a three-dimensional model of the plastic particle building block using a Bezier curve adaptive subdivision algorithm; Using finite element analysis algorithms, the thermal-mechanical coupling effect, stress distribution and material flow of the model during the injection molding process are dynamically simulated; Combined with the results of finite element analysis, numerical simulation algorithms are used to dynamically simulate and optimize the material fluidity, curing time, temperature changes and other process parameters during the injection molding process; Based on the finite element analysis and numerical simulation results, the optimal injection molding parameters are generated.

4. A precision CNC injection molding system according to claim 3, characterized in that: The construction process of the finite element analysis algorithm includes the following steps: Based on the three-dimensional model of plastic particle building blocks, the model is meshed using adaptive meshing technology; According to the physical and mechanical properties of the plastic particles, corresponding material parameters are set; the material parameters include density, elastic modulus, thermal conductivity, viscosity and specific heat capacity of the plastic particles; Combined with the injection molding process requirements of plastic particle building blocks, reasonable boundary conditions and initial conditions are set. The boundary conditions include injection pressure, mold temperature and flow channel design, and the initial conditions include the setting of initial temperature field and velocity field. According to the thermal-mechanical coupling effect and nonlinear behavior of materials involved in the injection molding process, the implicit finite element method is used, and the multi-step iteration technology is combined to solve the complex injection molding process; After solving the problem through simulation, the calculation results are analyzed using post-processing tools to generate data on the temperature distribution, stress-strain distribution, and flow path of the plastic particles during the injection molding process.

5. A precision CNC injection molding system according to claim 1, characterized in that: The feedback optimization module realizes real-time update of optimization algorithm model parameters of large-scale injection molding process data based on the cloud database, and automatically iterates and optimizes the control solution according to the big data analysis results in the production line.

Images