Artificial intelligence-based die-casting machine injection speed control method and system

Abstract

The present invention discloses an artificial intelligence-based die-casting machine injection speed control method and system, which relate to the technical field of die-casting process control, including collecting multi-dimensional data; using a deep learning model to extract spatiotemporal features of the multi-dimensional data, and analyzing the filling state, flow path and ideal injection speed of the molten metal based on the spatiotemporal features; identifying the thermal field imbalance area inside the mold according to the filling state and flow path of the molten metal and in combination with the ideal injection speed; adjusting the operating state of the die-casting machine in real time according to the thermal field imbalance area inside the mold, and obtaining the mold temperature distribution and the flow speed of the molten metal; adjusting the injection speed of the die-casting machine in real time according to the mold temperature distribution and the flow speed of the molten metal; the present invention extracts the spatiotemporal features of the multi-dimensional data through a deep learning model, accurately analyzes the filling state and flow path of the molten metal, and adjusts the injection speed in real time, thereby significantly improving the quality consistency of die-castings.

Description

Technical Field

[0001] The present invention relates to the technical field of die-casting process control, and in particular to an artificial intelligence-based die-casting machine injection speed control method and system. Background Art

[0002] In modern manufacturing, die-casting technology, as an efficient metal forming process, has been widely used in industrial fields such as automobiles, aerospace, and electronics. As the market's requirements for product quality continue to increase, the operating accuracy and control complexity of die-casting machines are also increasing. Traditional die-casting machine control systems mainly rely on preset speed curves and empirical parameters. Although these methods can guarantee production efficiency to a certain extent, they are unable to cope with complex working conditions. In recent years, with the development of artificial intelligence (AI) technology, especially the advancement of deep learning algorithms, new ideas and technical means have been provided to solve the problems in the traditional die-casting process.

[0003] However, the existing rule-based traditional die-casting machine control systems have two obvious shortcomings: first, they are difficult to adapt to the complex and changeable thermal field distribution inside the mold in real time, which may lead to local overheating or insufficient cooling, affecting the quality of the final product; second, due to the lack of accurate prediction of the metal liquid filling state and flow path, the traditional method often cannot dynamically adjust the injection speed to optimize the filling effect, resulting in incomplete filling or defects such as pores. The above problems limit the quality stability and manufacturing accuracy of die-casting parts, and thus affect the reliability and economic benefits of the entire production process. Summary of the invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides an artificial intelligence-based die-casting machine injection speed control method to solve the problem that traditional methods are difficult to adapt to thermal field changes in real time and accurately predict the metal liquid filling state.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides a die-casting machine injection speed control method based on artificial intelligence, which includes collecting multidimensional data; using a deep learning model to extract spatiotemporal features of the multidimensional data, and analyzing the filling state, flow path and ideal injection speed of the molten metal based on the spatiotemporal features; identifying the thermal field imbalance area inside the mold according to the filling state and flow path of the molten metal and in combination with the ideal injection speed; adjusting the operating state of the die-casting machine in real time according to the thermal field imbalance area inside the mold, and obtaining the mold temperature distribution and the flow speed of the molten metal; adjusting the injection speed of the die-casting machine in real time according to the mold temperature distribution and the flow speed of the molten metal; the specific steps of identifying the thermal field imbalance area inside the mold according to the filling state and flow path of the molten metal and in combination with the ideal injection speed are as follows:

[0008] Using the mold geometry and filling status, the mold interior is divided into multiple zones, and the average temperature and temperature gradient of each zone are identified;

[0009] Based on the deviation between the average temperature of each area and the overall average temperature, the thermal field balance index is defined , and combined with the ideal injection speed, the influence of injection speed on thermal field distribution is evaluated, and the expression is:

[0010] ;

[0011] in, is the actual injection speed, is the ideal injection speed, It is an index of thermal field balance. is the start time of the injection process, is the end time of the injection process, is the influence of injection speed on thermal field distribution;

[0012] Define the thermal field balance impact tolerance based on historical thermal field balance requirements and historical temperature fluctuation range ;

[0013] when When , it is considered that the current area is slightly affected by the injection velocity deviation and the thermal field distribution is balanced;

[0014] when When , it is considered that the current area is seriously affected by the injection velocity deviation and the thermal field distribution is uneven;

[0015] According to the thermal field imbalance area inside the mold, the operating state of the die-casting machine is adjusted in real time, and the mold temperature distribution and the flow rate of the molten metal are obtained. The specific steps are as follows:

[0016] When the heat field is unevenly distributed, record the residence time of the molten metal in each area, and calculate the average residence time based on the residence time of the molten metal in all areas;

[0017] Identify areas where the metal liquid residence time is greater than the average residence time as high-temperature areas, and optimize the layout of cooling channels to increase the cooling medium flow rate;

[0018] Identify areas where the metal liquid residence time is less than the average residence time as cold spots and reduce the number of cooling channels;

[0019] PID is used to monitor in real time the temperature changes in various areas of the mold and the flow rate of the molten metal during the injection process of the die-casting machine.

[0020] As a preferred solution of the artificial intelligence-based die-casting machine injection speed control method described in the present invention, the multidimensional data includes voiceprint signals, mold pressure changes, flow rate information, and temperature distribution data of the mold and molten metal.

[0021] As a preferred solution of the artificial intelligence-based die-casting machine injection speed control method of the present invention, wherein: the time-space domain feature extraction of multidimensional data using a deep learning model is performed, and the specific steps are as follows:

[0022] Normalize the multidimensional data, bind the timestamp and spatial coordinates of each multidimensional data, and obtain standardized spatiotemporal data;

[0023] Based on the standardized spatiotemporal data, LSTM is used to extract the frequency change characteristics and amplitude change characteristics of the voiceprint signal;

[0024] Based on the standardized spatiotemporal data, Transformer is used to extract the pressure change characteristics, flow velocity change characteristics, and the time series change characteristics of the mold and molten metal temperature distribution.

[0025] As a preferred solution of the artificial intelligence-based die-casting machine injection speed control method of the present invention, wherein: the filling state, flow path and ideal injection speed of the molten metal are analyzed based on the time-space domain characteristics, and the specific steps are as follows:

[0026] The frequency change characteristics and amplitude change characteristics of the voiceprint signal extracted by LSTM are fused with the pressure change characteristics, flow velocity change characteristics, and time series change characteristics of the mold and molten metal temperature distribution extracted by Transformer to generate comprehensive spatiotemporal characteristics. The expression is:

[0027] ;

[0028] in, It is a comprehensive spatial and temporal feature. is the i-th spatiotemporal feature, is the weight matrix of the i-th spatiotemporal feature, is the attention weight of the i-th spatiotemporal feature, is the numerical stability factor, n is the total number of spatiotemporal features, and i is the index variable of the spatiotemporal features;

[0029] Based on comprehensive spatiotemporal features , using 3D-CNN to analyze the real-time filling status of the molten metal in the mold; the filling status includes the filling position, unfilled area and filling integrity;

[0030] By utilizing the spatial relationship in the spatiotemporal features and combining the mold geometry, GNN is used to identify the flow path of the molten metal in the mold.

[0031] According to the filling state and flow path, combined with the pressure distribution in the mold and the temperature of the molten metal, the ideal injection speed is predicted as follows:

[0032] ;

[0033] in, It's time The ideal injection speed is is the real-time filling state matrix predicted by 3D-CNN, is the metal liquid flow path predicted by GNN, is the real-time pressure inside the mold, is the real-time temperature inside the mold, is the starting time point of the injection process, is the current time point, It represents the rate of change of the metal liquid flow velocity in space. It is a comprehensive conditional feature vector integrated by multi-layer MLP. is the rate of change of the real-time temperature in the mold, Represents the element-by-element multiplication of matrices.

[0034] As a preferred solution of the die-casting machine injection speed control method based on artificial intelligence of the present invention, wherein: the die-casting machine injection speed is adjusted in real time according to the mold temperature distribution and the flow speed of the molten metal, and the specific steps are as follows:

[0035] When the temperature of certain areas in the mold increases, reduce the injection speed to reduce the kinetic energy of the molten metal;

[0036] When the temperature of certain areas in the mold decreases, increase the injection speed to increase the fluidity of the molten metal;

[0037] When the flow rate of the molten metal decreases, increase the injection speed;

[0038] When the metal liquid flow speed is too high, reduce the injection speed.

[0039] In the second aspect, the present invention provides an artificial intelligence-based die-casting machine injection speed control system, including a data acquisition module, a feature extraction module, a region identification module, an operation state adjustment module and an injection speed adjustment module; a data acquisition module is used to collect multidimensional data; a feature extraction module is used to use a deep learning model to extract spatiotemporal features of multidimensional data, and analyze the filling state, flow path and ideal injection speed of the molten metal based on the spatiotemporal features; a region identification module is used to identify the thermal field imbalance area inside the mold according to the filling state and flow path of the molten metal and in combination with the ideal injection speed; an operation state adjustment module is used to adjust the operation state of the die-casting machine in real time according to the thermal field imbalance area inside the mold, and obtain the mold temperature distribution and the flow speed of the molten metal; an injection speed adjustment module is used to adjust the injection speed of the die-casting machine in real time according to the mold temperature distribution and the flow speed of the molten metal.

[0040] In a third aspect, the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, any step of the artificial intelligence-based die-casting machine injection speed control method as described in the first aspect of the present invention is implemented.

[0041] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, any step of the artificial intelligence-based die-casting machine injection speed control method as described in the first aspect of the present invention is implemented.

[0042] The beneficial effects of the present invention are as follows: by extracting the spatiotemporal features of multidimensional data through a deep learning model, the filling state and flow path of the molten metal are accurately analyzed, and the injection speed is adjusted in real time, which significantly improves the quality consistency of the die-casting parts. In addition, by identifying and compensating for the thermal field imbalance in the mold, the die-casting process is further optimized, ensuring the ideal thermal field distribution, and improving production efficiency and product quality. These two core steps together achieve smarter and more flexible die-casting control, solving the problem that traditional methods are difficult to adapt to thermal field changes and accurately predict filling states. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0044] Figure 1This is a flow chart of the artificial intelligence-based die-casting machine injection speed control method in Example 1.

[0045] Figure 2 This is a schematic diagram of the injection speed control of the die-casting machine based on artificial intelligence in Example 1. DETAILED DESCRIPTION

[0046] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0047] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0048] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0049] Example 1, reference Figure 1 and Figure 2 , which is the first embodiment of the present invention, provides a die-casting machine injection speed control method based on artificial intelligence, comprising the following steps:

[0050] S1: Collect multidimensional data;

[0051] S1.1: Multidimensional data includes voiceprint signals, mold pressure changes, flow rate information, and mold and molten metal temperature distribution data.

[0052] Furthermore, the microphone array installed around the die-casting machine is used to collect voiceprint signal data during operation, record the working sound of the servo motor, the airflow noise of the heat exchange system, and the weak sound signals during the molten metal filling process; secondly, the pressure sensor and flow rate sensor in the mold cavity are used to collect the pressure changes and flow rate information of the molten metal in the mold cavity in real time; finally, the temperature distribution data of the mold surface and the molten metal flow channel is obtained through the thermal imaging sensor.

[0053] S2: Use deep learning models to extract spatiotemporal features from multidimensional data, and analyze the filling state, flow path, and ideal injection speed of the molten metal based on the spatiotemporal features.

[0054] S2.1: Normalize the multidimensional data and bind each multidimensional data to a timestamp and spatial coordinates to obtain standardized spatiotemporal data.

[0055] It should be noted that first, multidimensional data from the die-casting process is collected, including but not limited to voiceprint signals, mold pressure changes, flow rate information, and temperature distribution data of the mold and molten metal. In order to ensure that data of different dimensions can be effectively integrated in the same model and avoid certain features being dominated or ignored due to excessively large or small numerical ranges, these multidimensional data are normalized, that is, the original data is adjusted to a preset standard range (such as the [0, 1] interval) through linear transformation or other appropriate methods. Next, a corresponding timestamp is added to each set of normalized data to record the data collection time, and the spatial coordinates are bound to identify the specific location of the data inside the mold.

[0056] S2.2: Based on the standardized spatiotemporal data, LSTM is used to extract the frequency change characteristics and amplitude change characteristics of the voiceprint signal.

[0057] It should be noted that the long short-term memory network (LSTM), a variant of the recurrent neural network (RNN) that is particularly suitable for processing sequence data, is used to analyze the standardized voiceprint signal. LSTM can capture long-term dependencies and is very suitable for mining hidden patterns from time series data. The voiceprint signal in the standardized spatiotemporal data is input into the LSTM model, and the model automatically learns and extracts the frequency change characteristics and amplitude change characteristics of the voiceprint signal through the internal memory unit. For example, the vibration generated when the molten metal fills the mold will be converted into a specific voiceprint signal. LSTM can identify the change trend of these signals over time, thereby helping to understand the dynamic behavior of the molten metal during the flow process.

[0058] S2.3: Based on the standardized spatiotemporal data, Transformer is used to extract the pressure change characteristics and flow velocity change characteristics, as well as the time series change characteristics of the mold and molten metal temperature distribution.

[0059] It should be noted that the Transformer architecture is a powerful tool that was originally designed for natural language processing tasks but is also applicable to a variety of sequence modeling problems. For features such as pressure changes, flow rate changes, and temperature distribution in standardized spatiotemporal data, the Transformer can efficiently handle long-distance dependencies through a self-attention mechanism and calculate the associations between different time points in parallel. Specifically, the Transformer model is used to parse these time series data to extract pressure change features, flow rate change features, and time series change features of the mold and molten metal temperature distribution. For example, in the die-casting process, as the molten metal gradually fills the mold, the internal pressure and temperature will undergo complex changes; the Transformer can effectively capture these change patterns and provide deep insights into the filling state and heat exchange process, which is crucial for optimizing injection speed control.

[0060] S2.4: The frequency change characteristics and amplitude change characteristics of the voiceprint signal extracted by LSTM are integrated with the pressure change characteristics, flow velocity change characteristics, and time series change characteristics of the mold and molten metal temperature distribution extracted by Transformer to generate comprehensive spatiotemporal characteristics. The expression is:

[0061] ;

[0062] in, It is a comprehensive spatial and temporal feature. is the i-th spatiotemporal feature, is the weight matrix of the i-th spatiotemporal feature, is the attention weight of the i-th spatiotemporal feature, is a numerical stability factor, n is the total number of spatiotemporal features, and i is the index variable of the spatiotemporal features.

[0063] It should be noted that the frequency change features and amplitude change features of the voiceprint signal extracted by the Long Short-Term Memory (LSTM) network are fused with the pressure change features, flow rate change features, and time series change features of the mold and molten metal temperature distribution extracted by the Transformer architecture. These features are first processed by nonlinear transformation and attention mechanism to adjust the importance of different features. Subsequently, the processed features are combined and processed by normalization and activation function to generate comprehensive spatiotemporal features. This comprehensive spatiotemporal feature can fully reflect the various dynamic changes in the die-casting process and provide a detailed information basis for subsequent analysis. This process ensures that spatiotemporal data from different sources can be effectively integrated, thereby more accurately capturing the complex behavior patterns in the die-casting process.

[0064] S2.5: Based on comprehensive spatiotemporal features , 3D-CNN is used to analyze the real-time filling status of the molten metal in the mold; the filling status includes the filling position, unfilled area and filling completeness.

[0065] It should be noted that in this process, the comprehensive spatiotemporal features F are input into the 3D-CNN, which captures the complex patterns in the spatial and temporal dimensions of the metal liquid filling process through multi-layer convolution operations. 3D-CNN can identify and extract detailed information about the metal liquid filling position, unfilled areas, and filling integrity, thereby providing a high-resolution image of the distribution of the metal liquid in the mold. This makes it possible to accurately monitor the flow path and filling progress of the metal liquid, ensuring the efficiency of the die casting process and the consistency of product quality. Finally, through real-time analysis of the filling state, key data support is provided for optimizing the injection speed control.

[0066] S2.6: Using the spatial relationship in the spatiotemporal domain features and combining the mold geometry, GNN is used to identify the flow path of the molten metal in the mold.

[0067] It should be noted that in this process, the spatial information of the spatiotemporal features is combined with the geometric structure of the mold. By analyzing this graph model through GNN, the transfer mode and interaction between different positions of the molten metal inside the mold can be captured. GNN learns and predicts how the molten metal flows from the inlet to each area based on the connection relationship between the nodes and the weights on the edges, and finally determines the detailed flow path. This method not only takes into account the time series changes of the molten metal flow, but also fully considers the spatial layout inside the mold, thereby providing accurate flow path information for optimizing the injection speed control.

[0068] S2.7: According to the filling state and flow path, combined with the pressure distribution in the mold and the temperature of the molten metal, the ideal injection speed is predicted. The expression is:

[0069] ;

[0070] in, It's time The ideal injection speed is is the real-time filling state matrix predicted by 3D-CNN, is the metal liquid flow path predicted by GNN, is the real-time pressure inside the mold, is the real-time temperature inside the mold, is the starting time point of the injection process, is the current time point, It represents the rate of change of the metal liquid flow velocity in space. It is a comprehensive conditional feature vector integrated by multi-layer MLP. is the rate of change of the real-time temperature in the mold, Represents the element-by-element multiplication of matrices.

[0071] It should be noted that the ideal injection speed is predicted by a mathematical expression based on the real-time filling state matrix predicted by 3D-CNN and the molten metal flow path predicted by Graph Neural Network (GNN), combined with the real-time pressure in the mold and the molten metal temperature. This process first integrates the filling state, flow path, pressure and temperature information to reflect the impact of these factors on the injection speed. Next, the cumulative effect from the start of injection to the current time point is calculated to ensure that the dynamic changes in the entire process are taken into account. In order to ensure numerical stability and take into account the spatial changes in the flow velocity and temperature of the molten metal, the expression is also appropriately adjusted. Finally, the calculation results are processed by the activation function to obtain the current ideal injection speed.

[0072] S3: Identify the thermal field imbalance area inside the mold based on the filling state and flow path of the molten metal and the ideal injection speed.

[0073] S3.1: Using the mold geometry and filling status, divide the mold interior into multiple zones and identify the average temperature and temperature gradient in each zone.

[0074] It should be noted that in this process, firstly, a reasonable division scheme is determined according to the geometric structure of the mold and the filling state of the molten metal to ensure that each area can represent specific thermal field characteristics. Then, in each divided area, the average temperature in the area is calculated, and the temperature gradient, that is, the rate of change of temperature in space, is evaluated. In this way, not only the temperature distribution of each area can be identified, but also the temperature change trend and imbalance can be accurately captured.

[0075] S3.2: Define thermal field balance index based on the deviation of the average temperature of each area from the overall average temperature , and combined with the ideal injection speed, the influence of injection speed on thermal field distribution is evaluated, and the expression is:

[0076] ;

[0077] in, is the actual injection speed, is the ideal injection speed, It is an index of thermal field balance. is the start time of the injection process, is the end time of the injection process, It is the influence of injection speed on thermal field distribution.

[0078] It should be noted that in this process, the deviation of the average temperature in each area relative to the average temperature of the entire mold is first calculated to define the thermal field balance index. Then, the difference between the actual injection speed and the ideal injection speed is compared, and combined with the thermal field balance index, an analysis is performed from the start to the end of the injection process. This analysis reflects the degree of influence of the injection speed deviation on the thermal field distribution. Finally, by quantifying this degree of influence, the impact of the injection speed change on the thermal field balance inside the mold can be accurately measured, providing a scientific basis for optimizing the injection speed control strategy.

[0079] S3.3: Define the thermal field balance impact tolerance based on historical thermal field balance requirements and historical temperature fluctuation range .

[0080] when When , it is considered that the current area is slightly affected by the injection velocity deviation and the thermal field distribution is balanced.

[0081] For example, =0.05 =0.1, it is considered that the area is slightly affected by the injection velocity deviation and the thermal field distribution remains balanced.

[0082] when When , it is considered that the current area is seriously affected by the injection velocity deviation and the thermal field distribution is uneven.

[0083] For example, =0.15 =0.1, it is considered that the area is seriously affected by the injection velocity deviation, resulting in uneven thermal field distribution.

[0084] S4: According to the thermal field imbalance area inside the mold, the operating state of the die-casting machine is adjusted in real time, and the mold temperature distribution and the flow rate of the molten metal are obtained.

[0085] S4.1: When the heat field is unevenly distributed, record the residence time of the molten metal in each area, and calculate the average residence time based on the residence time of the molten metal in all areas.

[0086] It should be noted that when the heat field distribution in the mold is detected to be uneven, the residence time of the molten metal in each divided area will be recorded. By analyzing these residence time data and summarizing the residence time of all areas, an average residence time value is obtained.

[0087] S4.2: Identify the area where the metal liquid residence time is greater than the average residence time as a high-temperature area, and optimize the layout of the cooling channel to increase the cooling medium flow rate.

[0088] It should be noted that areas where the metal liquid stays longer than the average residence time are identified as high-temperature areas. For these high-temperature areas, measures are taken to optimize the layout of the cooling channels and increase the flow rate of the cooling medium accordingly. This step is intended to accelerate heat dissipation, ensure that the temperature of these areas drops rapidly, and maintain the balance of the thermal field inside the entire mold.

[0089] S4.3: Identify areas where the metal liquid residence time is less than the average residence time as cold spot areas and reduce the number of cooling channels.

[0090] It should be noted that areas where the residence time of the molten metal is lower than the average residence time are marked as cold spots. In order to improve the temperature conditions in these areas, the number of cooling channels is appropriately reduced. This step helps to avoid overcooling and ensure that the cold spot area can reach the ideal temperature level, thereby promoting the uniform distribution of the overall thermal field.

[0091] S4.5: Use PID to monitor in real time the temperature changes in various areas of the mold and the flow rate of the molten metal during the injection process of the die-casting machine.

[0092] It should be noted that during the entire injection process, a proportional integral derivative (PID) controller is used to monitor the temperature changes in various areas of the mold and the flow rate of the molten metal in real time. Through continuous data collection and analysis, the PID controller can respond to any changes in temperature or flow rate in a timely manner, ensuring that the conditions in the die-casting process are always maintained in the optimal state to achieve efficient and stable production.

[0093] S5: Adjust the injection speed of the die-casting machine in real time according to the mold temperature distribution and the flow speed of the molten metal.

[0094] S5.1: When the temperature in certain areas of the mold increases, reduce the injection speed to reduce the kinetic energy of the molten metal.

[0095] It should be noted that this process helps prevent quality problems caused by excessive temperature in specific areas, ensures a more uniform heat field distribution, and maintains the stability of the entire die-casting process.

[0096] S5.2: When the temperature in certain areas of the mold decreases, increase the injection speed to increase the fluidity of the molten metal.

[0097] It should be noted that this process is designed to ensure that the low-temperature area can be fully filled to avoid incomplete filling or defects caused by too low temperature, thereby optimizing the molding quality of the product.

[0098] S5.3: When the metal liquid flow rate decreases, increase the injection speed.

[0099] It should be noted that this process ensures that even when flow is obstructed, good filling effects are maintained and voids or other underfilling problems are avoided.

[0100] S5.4: When the metal liquid flow rate is too high, reduce the injection speed.

[0101] It should be noted that this process prevents unstable filling caused by excessive flow rate, such as splashing or overfilling, ensures a smooth die-casting process and improves product quality.

[0102] The present embodiment also provides an artificial intelligence-based die-casting machine injection speed control system, comprising: a data acquisition module, a feature extraction module, a region identification module, an operation state adjustment module and an injection speed adjustment module; the data acquisition module is used to collect multidimensional data; the feature extraction module is used to use a deep learning model to extract spatiotemporal features of the multidimensional data, and analyze the filling state, flow path and ideal injection speed of the molten metal based on the spatiotemporal features; the region identification module is used to identify the thermal field imbalance area inside the mold according to the filling state and flow path of the molten metal and in combination with the ideal injection speed; the operation state adjustment module is used to adjust the operation state of the die-casting machine in real time according to the thermal field imbalance area inside the mold, and obtain the mold temperature distribution and the flow speed of the molten metal; the injection speed adjustment module is used to adjust the injection speed of the die-casting machine in real time according to the mold temperature distribution and the flow speed of the molten metal.

[0103] This embodiment also provides a computer device, which is suitable for the case of an artificial intelligence-based die-casting machine injection speed control method, including: a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute computer executable instructions to implement the artificial intelligence-based die-casting machine injection speed control method proposed in the above embodiment.

[0104] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covered on the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0105] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by the processor, the method for controlling the injection speed of a die-casting machine based on artificial intelligence as proposed in the above embodiment is implemented; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as a static random access memory (SRAM), an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic storage device, a flash memory, a disk or an optical disk.

[0106] In summary, the present invention uses: deep learning model to extract spatiotemporal features of multidimensional data, accurately analyzes the filling state and flow path of the molten metal, and adjusts the injection speed in real time, significantly improving the quality consistency of die-castings. In addition, by identifying and compensating for the thermal field imbalance in the mold, the die-casting process is further optimized, ensuring the ideal thermal field distribution, and improving production efficiency and product quality. These two core steps together achieve smarter and more flexible die-casting control, solving the problem that traditional methods are difficult to adapt to thermal field changes and accurately predict filling states.

[0107] Example 2, referring to Table 1, is the second example of the present invention. To further verify the technical solution of the present invention, experimental simulation data of the injection speed control method of the die-casting machine based on artificial intelligence are provided.

[0108] In order to verify the effectiveness of the "artificial intelligence-based die-casting machine injection speed control method", an experiment was designed. Two different types of molds were selected for comparative testing: one was using the traditional method (called the control group), and the other was applying the method of the present invention (called the experimental group). Each mold was subjected to 10 independent die-casting processes to ensure the reliability and statistical significance of the data. In order to increase the reliability of the experiment, three different molds were set up under each condition (i.e., control groups A, B, C and experimental groups A, B, C), for a total of six molds.

[0109] The method of the present invention: First, the collected multidimensional data is normalized, and timestamps and spatial coordinates are added to each multidimensional data to generate standardized spatiotemporal data. The long short-term memory network (LSTM) is used to extract the frequency change characteristics and amplitude change characteristics of the voiceprint signal, and the Transformer architecture is used to extract the time series characteristics of pressure, flow rate and temperature distribution. These features are fused to generate a comprehensive spatiotemporal feature F, which reflects the filling state of the molten metal, the flow path, and the pressure and temperature changes in the mold.

[0110] Secondly, 3D-CNN analyzes the real-time filling status of the molten metal in the mold, while the graph neural network (GNN) identifies the specific flow path. Based on these analysis results, the ideal injection speed is predicted and adjusted to optimize the die casting process.

[0111] Next, in order to evaluate the thermal field balance, the mold was divided into multiple areas, the average temperature and temperature gradient of each area were calculated, and the thermal field balance index was defined. According to the thermal field imbalance, the cooling channel layout and injection speed were adjusted in real time to ensure the efficiency and stability of the process.

[0112] Finally, during the entire experiment, the PID controller continuously monitors the temperature changes in various areas of the mold and the flow rate of the molten metal, responds and adjusts the injection speed in a timely manner to maintain the ideal heat field distribution and filling effect, ensuring the best production conditions.

[0113] Traditional method: A die casting machine control system based on fixed parameters and preset rules. This method relies on pre-set pressure, speed and temperature parameters, as well as fixed cooling strategies, and lacks the ability to adjust and optimize in real time.

[0114] The details are shown in Table 1 below:

[0115] Table 1 Comparison of experimental data of die casting machine injection speed control method

[0116]

[0117] Through the data analysis of the above table, it can be clearly seen that the "artificial intelligence-based die-casting machine injection speed control method" is significantly better than the traditional control method in multiple key performance indicators. The present invention realizes the refined management and real-time optimization of the die-casting process by introducing deep learning models and intelligent algorithms. For example, the average defect rate of the experimental group was reduced from about 5.07% of the control group to 2.1%, the average filling time was shortened by nearly 2 seconds, the energy consumption was reduced by about 0.6 kWh, and the thermal field balance score was increased from 7.5 to 9.0, showing a more uniform temperature distribution and a more efficient production process. These improvements not only improve product quality and production efficiency, but also effectively save energy.

[0118] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A die-casting machine injection speed control method based on artificial intelligence, characterized in that: include, Collect multi-dimensional data; Use deep learning models to extract spatiotemporal features from multidimensional data, and analyze the filling state, flow path, and ideal injection speed of the molten metal based on spatiotemporal features; According to the filling state and flow path of the molten metal, combined with the ideal injection speed, the thermal field imbalance area inside the mold is identified; According to the thermal field imbalance area inside the mold, the operating status of the die-casting machine is adjusted in real time, and the mold temperature distribution and the flow rate of the molten metal are obtained; According to the temperature distribution of the mold and the flow speed of the molten metal, the injection speed of the die casting machine is adjusted in real time; according to the filling state and flow path of the molten metal, combined with the ideal injection speed, the thermal field imbalance area inside the mold is identified. The specific steps are as follows: Using the mold geometry and filling status, the mold interior is divided into multiple zones, and the average temperature and temperature gradient of each zone are identified; Based on the deviation of the average temperature of each area from the overall average temperature, the thermal field balance index is defined , and combined with the ideal injection speed, the influence of injection speed on thermal field distribution is evaluated, and the expression is: ； in, is the actual injection speed, is the ideal injection speed, It is an index of thermal field balance. is the start time of the injection process, is the end time of the injection process, is the influence of injection speed on thermal field distribution; Define the thermal field balance impact tolerance based on historical thermal field balance requirements and historical temperature fluctuation range ; when When , it is considered that the current area is slightly affected by the injection velocity deviation and the thermal field distribution is balanced; when When , it is considered that the current area is seriously affected by the injection velocity deviation and the thermal field distribution is uneven; According to the thermal field imbalance area inside the mold, the operating state of the die-casting machine is adjusted in real time, and the mold temperature distribution and the flow rate of the molten metal are obtained. The specific steps are as follows: When the heat field is unevenly distributed, record the residence time of the molten metal in each area, and calculate the average residence time based on the residence time of the molten metal in all areas; Identify areas where the metal liquid residence time is greater than the average residence time as high-temperature areas, and optimize the layout of cooling channels to increase the cooling medium flow rate; Identify areas where the metal liquid residence time is less than the average residence time as cold spots and reduce the number of cooling channels; PID is used to monitor in real time the temperature changes in various areas of the mold and the flow rate of the molten metal during the injection process of the die-casting machine.

2. The artificial intelligence-based die-casting machine injection speed control method according to claim 1, characterized in that: The multi-dimensional data includes voiceprint signals, mold pressure changes, flow rate information, and temperature distribution data of the mold and the molten metal.

3. The artificial intelligence-based die-casting machine injection speed control method according to claim 2, characterized in that: The specific steps of using the deep learning model to extract spatiotemporal features from multidimensional data are as follows: Normalize the multidimensional data, bind the timestamp and spatial coordinates of each multidimensional data, and obtain standardized spatiotemporal data; Based on the standardized spatiotemporal data, LSTM is used to extract the frequency change characteristics and amplitude change characteristics of the voiceprint signal; Based on the standardized spatiotemporal data, Transformer is used to extract the pressure change characteristics, flow velocity change characteristics, and the time series change characteristics of the mold and molten metal temperature distribution.

4. The artificial intelligence-based die-casting machine injection speed control method according to claim 3, characterized in that: The specific steps of analyzing the filling state, flow path and ideal injection speed of the molten metal based on the time-space domain characteristics are as follows: The frequency change characteristics and amplitude change characteristics of the voiceprint signal extracted by LSTM are fused with the pressure change characteristics, flow velocity change characteristics, and time series change characteristics of the mold and molten metal temperature distribution extracted by Transformer to generate comprehensive spatiotemporal characteristics. The expression is: ； in, It is a comprehensive spatial and temporal feature. is the i-th spatiotemporal feature, is the weight matrix of the i-th spatiotemporal feature, is the attention weight of the i-th spatiotemporal feature, is the numerical stability factor, n is the total number of spatiotemporal features, and i is the index variable of the spatiotemporal features; Based on comprehensive spatiotemporal features , using 3D-CNN to analyze the real-time filling status of the molten metal in the mold; the filling status includes the filling position, unfilled area and filling integrity; By utilizing the spatial relationship in the spatiotemporal features and combining the mold geometry, GNN is used to identify the flow path of the molten metal in the mold. According to the filling state and flow path, combined with the pressure distribution in the mold and the temperature of the molten metal, the ideal injection speed is predicted as follows: ； in, It's time The ideal injection speed is is the real-time filling state matrix predicted by 3D-CNN, is the metal liquid flow path predicted by GNN, is the real-time pressure inside the mold, is the real-time temperature inside the mold, is the starting time point of the injection process, is the current time point, It represents the rate of change of the metal liquid flow velocity in space. It is a comprehensive conditional feature vector integrated by multi-layer MLP. is the rate of change of the real-time temperature in the mold, Represents the element-by-element multiplication of matrices.

5. The artificial intelligence-based die-casting machine injection speed control method according to claim 4, characterized in that: The specific steps of adjusting the injection speed of the die casting machine in real time according to the mold temperature distribution and the flow speed of the molten metal are as follows: When the temperature of certain areas in the mold increases, reduce the injection speed to reduce the kinetic energy of the molten metal; When the temperature of certain areas in the mold decreases, increase the injection speed to increase the fluidity of the molten metal; When the flow rate of the molten metal decreases, increase the injection speed; When the metal liquid flow speed is too high, reduce the injection speed.

6. An artificial intelligence-based die-casting machine injection speed control system, based on the artificial intelligence-based die-casting machine injection speed control method according to any one of claims 1 to 5, characterized in that: It includes a data acquisition module, a feature extraction module, a region recognition module, an operation state adjustment module and an injection speed adjustment module; A data acquisition module, used for collecting multi-dimensional data; The feature extraction module is used to extract spatiotemporal features of multidimensional data using a deep learning model, and analyze the filling state, flow path, and ideal injection speed of the molten metal based on the spatiotemporal features; The area recognition module is used to identify the thermal field imbalance area inside the mold based on the filling state and flow path of the molten metal and the ideal injection speed; The operating state adjustment module is used to adjust the operating state of the die-casting machine in real time according to the thermal field imbalance area inside the mold, and obtain the mold temperature distribution and the flow speed of the molten metal; The injection speed adjustment module is used to adjust the injection speed of the die-casting machine in real time according to the mold temperature distribution and the flow speed of the molten metal.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the artificial intelligence-based die-casting machine injection speed control method described in any one of claims 1 to 5 are implemented.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the artificial intelligence-based die-casting machine injection speed control method according to any one of claims 1 to 5 are implemented.

Images