High-entropy eutectic ceramic powder material preparation device and control system thereof

Abstract

The application discloses a high-entropy eutectic ceramic powder material preparation equipment and a control system thereof and relates to the technical field of ceramic powder preparation. The equipment comprises a feeding control module. The feeding control module is used for performing metering input of multiple-component raw materials through a feeding assembly, obtaining raw material weight data to obtain an initial weight set, identifying raw material batch information to obtain an initial information set, corresponding the initial information set to the initial weight set, setting a central dynamic control system, and obtaining initial mixing information through the central dynamic control system. The application realizes linkage of RFID batch identification and an intelligent metering feeder, realizes real-time acquisition of raw material purity and density data, automatically performs dynamic weight compensation and purity conversion when raw material batch parameters deviate from preset values, and ensures that actual effective components of multiple-component raw materials meet a preset molar ratio, so that the risk of component segregation caused by raw material fluctuation is reduced from the source.

Description

Technical Field

[0001] This invention relates to the field of ceramic powder preparation technology, specifically to a high-entropy eutectic ceramic powder material preparation equipment and its control system. Background Technology

[0002] High-entropy eutectic ceramic powder is a ceramic material composed of multiple elements that forms a "eutectic" phase structure under specific temperature and composition conditions. High-entropy materials refer to alloys or compounds composed of five or more main elements in near equimolar ratios in a crystal, possessing a high "formulation entropy." Eutectic refers to a mixture of a specific composition on a phase diagram that simultaneously sublimates / melts into two or more phases at a specific temperature, possessing a minimum eutectic temperature and a finer two- or multi-phase symbiotic structure. Through the high mixing degree of multi-components, the single-crystal type and single-phase stability in traditional ceramic systems can be broken, resulting in finer grains and complex multi-phase microstructures, thereby improving mechanical properties, thermal stability, wear resistance, thermal shock resistance, and fracture toughness. In the preparation process, multi-component raw materials are usually thoroughly mixed.

[0003] A dielectric ceramic powder and its preparation method are disclosed in patent publication number CN118724591A. The dielectric constant of the dielectric ceramic powder is adjustable, and it has excellent high-temperature stability. When working in a high-temperature environment, the products prepared from the dielectric ceramic powder can maintain good dielectric properties. The products prepared from the dielectric ceramic powder have low dielectric loss and high energy storage density. The dielectric ceramic powder can be sintered to obtain small products, and the miniaturized products still have high breakdown field strength, good energy storage performance, and dielectric property stability.

[0004] When the above-mentioned and similar technical solutions are used, in the powder mixing process of existing ceramic powder preparation, when there are a large number of multi-component high-entropy ceramics, such as five or more metal oxides or carbides present at the same time, the density differences between different metal oxides or carbides are large, such as the density difference between HfO2 and Al2O2 reaching 5 g / cm³. 3 Furthermore, the particle size is uneven, which can easily lead to segregation. In addition, mechanical mixing can easily introduce impurities from grinding balls, affecting the eutectic performance. Traditional mixing equipment relies on manual adjustment and lacks real-time feedback, resulting in large fluctuations in mixing uniformity, which further affects the final eutectic structure performance. Summary of the Invention

[0005] The purpose of this invention is to provide a high-entropy eutectic ceramic powder material preparation device and its control system to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a high-entropy eutectic ceramic powder material preparation control system, comprising:

[0007] Feeding control module: The feeding component executes the metering input of multi-component raw materials, obtains the raw material weight data to obtain the initial weight set, and identifies the raw material batch information to obtain the initial information set. The initial information set corresponds to the initial weight set. The central dynamic control system is set up, and the initial mixing information is obtained through the central dynamic control system. The initial mixing information includes the metering information of multi-component raw materials to obtain the initial mixing item. Based on the initial mixing item, the feeding component adds raw materials to obtain the initial feeding item.

[0008] Raw material mixing module: Set up a combined mixing device, and perform first mixing and second mixing through the combined mixing device. Set a first initial value and a second initial value based on the first mixing and the second mixing, respectively. Perform first mixing with the first initial value and second mixing with the second initial value to obtain the combined mixing item;

[0009] Hybrid computing module: Based on the combined hybrid terms, it obtains the hybrid real-time state information to obtain the real-time information terms. Using the real-time information terms as training data, it trains the model through the central dynamic control system, runs the deep reinforcement learning model, and outputs the hybrid parameter adjustment instructions to obtain the output adjustment terms.

[0010] The reverse adjustment module adjusts the initial frequency value, initial field strength value, and initial mixing term based on the output adjustment term, thereby realizing the preparation of multi-component high-entropy eutectic ceramic powder. To address the problems of uneven mixing and component segregation, a deep learning model is used to analyze sensor data in real time and dynamically adjust the mixing parameters to achieve a self-optimized mixing process. Closed-loop control is used to suppress segregation caused by density differences and ensure low molar ratio error of high-entropy components.

[0011] Furthermore, the method for obtaining the initial feed item includes:

[0012] The weight data of multiple raw materials are acquired in real time by a weight sensor to obtain an initial weight set, which includes at least one initial weight item consisting of the weight of a raw material.

[0013] By combining RFID tag identification components, batch information of multi-component raw materials is identified, including purity information and density information, to obtain an initial information set. The initial information set includes at least one initial information item consisting of raw material purity information and density information.

[0014] Based on the initial mixing item, the composition ratio information of the initial weight item and the initial information item is obtained through data acquisition, and then the raw materials are added through the feeding component to obtain the initial feeding item.

[0015] Furthermore, the combined mixing device includes a high-energy ultrasonic and electromagnetic mixing reaction component, and the method for obtaining the combined mixing item includes:

[0016] Set an initial frequency value and an initial field strength value, using the initial frequency value as the first initial value and the initial field strength value as the second initial value;

[0017] Ultrasonic oscillation is performed at the initial frequency value, and electromagnetic stirring is performed at the initial field strength value, thereby obtaining a combined mixture term.

[0018] Furthermore, the method for obtaining the initial frequency value and the initial field strength value includes:

[0019] The working mixing degree of the high-energy ultrasound and electromagnetic mixing reaction components was obtained separately, including the ultrasonic oscillation mixing degree and the electromagnetic stirring mixing degree.

[0020] Based on the mapping relationship between the initial mixing term and the ultrasonic oscillation mixing degree and the electromagnetic stirring mixing degree, the oscillation mixing frequency set and the stirring mixing field strength set are obtained. The oscillation mixing frequency set includes at least one oscillation mixing frequency term, and the stirring mixing field strength set includes at least one stirring mixing field strength term. The oscillation mixing frequency term and the stirring mixing field strength term correspond to each other, and at least one set of adjustment information is obtained.

[0021] The oscillation mixing frequency term and the stirring mixing field strength term corresponding to the adjustment information group are used as the initial frequency value and the initial field strength value, respectively.

[0022] Furthermore, the method for obtaining the initial frequency value and the initial field strength value also includes:

[0023] Energy consumption data of the oscillation mixing frequency set and the stirring mixing field strength set are obtained respectively. The energy consumption information corresponding to the adjustment information group is sorted in descending order to obtain the energy consumption ranking. The last information of the energy consumption ranking item is obtained as a control adjustment. Based on the control adjustment, the oscillation mixing frequency item and the stirring mixing field strength item are obtained respectively, and then the initial frequency value and the initial field strength value are obtained.

[0024] Furthermore, the hybrid real-time state information includes a uniformity index, and the method for obtaining the real-time information items includes:

[0025] The acquisition equipment is set up, including an in-situ laser-induced breakdown spectrometer. Based on the in-situ laser-induced breakdown spectrometer, the pulse laser index, including the wavelength index and the pulse width index, is set to obtain the laser adjustment term.

[0026] The incident angle is set based on the laser adjustment term. Based on the incident angle, the laser is injected into the mixed powder layer to break through the surface and generate plasma. The capture range and at least two capture points are set. The plasma emission spectrum is captured by a spectral sensor to obtain the capture information.

[0027] Based on the captured information item, the spectral intensity is converted into elemental concentration using the standard curve method, and the uniformity index is calculated to obtain the real-time information item.

[0028] Furthermore, the hybrid real-time status information also includes temperature gradient data, and the method for obtaining the real-time information items includes:

[0029] A three-dimensional temperature field scan is performed using an infrared thermal imaging array and an embedded thermocouple network. The number of infrared thermal imaging arrays is at least two. The surface temperature distribution of the combined hybrid device is scanned to obtain the scanned temperature item. Based on the scanned combined hybrid device with at least two embedded thermocouple networks embedded inside, the temperature of the core area is measured to obtain the measured temperature item.

[0030] Based on the positional comparison information of the scanned temperature item and the measured temperature item, a temperature gradient map is obtained, and then real-time information items are obtained.

[0031] Furthermore, the method for obtaining the output adjustment term includes:

[0032] Using the initial feeding item, the first initial value, the second initial value, and the real-time information item as input data, timestamp alignment and outlier cleaning are performed to obtain the processed dataset;

[0033] Based on the processed dataset, core features are extracted, feature vectors are mapped to the state space of reinforcement learning, the algorithm framework and network structure are set, the reinforcement learning decision mechanism is set, including the definition of action space and the design of reward function, and finally the parameter adjustment instructions are generated and output, thus obtaining the output adjustment term.

[0034] A high-entropy eutectic ceramic powder material preparation device, using the aforementioned high-entropy eutectic ceramic powder material preparation control system, includes:

[0035] The feeding assembly includes a weight sensing unit and a batch identification unit, used to perform metering input of multi-component raw materials;

[0036] A combined mixing device for first and second mixing of multi-component raw materials;

[0037] The data acquisition unit is used to acquire real-time mixing status information of multiple component raw materials;

[0038] The central dynamic control system is used to run deep reinforcement learning models and output mixed parameter adjustment commands.

[0039] Compared with the prior art, the beneficial effects of the present invention are:

[0040] This high-entropy eutectic ceramic powder material preparation equipment and its control system, through RFID batch identification and linkage with an intelligent metering feeder, acquires raw material purity and density data in real time. When the raw material batch parameters deviate from the preset values, the system automatically performs dynamic weight compensation and purity conversion to ensure that the actual effective components of the multi-component raw materials meet the preset molar ratio, thereby reducing the risk of component segregation caused by raw material fluctuations from the source. At the same time, it adopts a combination of high-energy ultrasound and electromagnetic mixing, based on a deep reinforcement learning model, to analyze in real time the uniformity index obtained by in-situ LIBS and the temperature gradient data obtained by infrared thermal imaging and thermocouple network, and dynamically adjust the mixing parameters such as ultrasonic frequency and electromagnetic field strength to form a closed-loop control, effectively overcoming the significant density differences between different components and improving the mixing uniformity. Attached Figure Description

[0041] Figure 1 This is a schematic diagram of the overall process of the present invention;

[0042] Figure 2 This is a schematic diagram of the initial material input item acquisition process of the present invention;

[0043] Figure 3 This is a schematic diagram illustrating the adjustment of the composition ratio information of the present invention;

[0044] Figure 4 This is a schematic diagram showing the relationship between the oscillation mixing frequency set and the stirring mixing field strength set of the present invention. Detailed Implementation

[0045] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0046] When there are many components in a multi-component high-entropy ceramic, such as five or more metal oxides or carbides, the significant density differences and varying particle sizes between the components can easily lead to segregation. If traditional mechanical mixing methods are used, heavy components tend to sink while light components float, resulting in uneven mixing. Furthermore, mechanical mixing carries the risk of introducing impurities from grinding balls. Even minute impurities can negatively impact the eutectic properties of the final material during high-entropy ceramic preparation. During high-speed grinding, grinding balls inevitably wear down, and these worn particles become part of the powder, difficult to remove completely and reducing the purity of the ceramic material. On the other hand, traditional mixing equipment relies on manual adjustment and lacks real-time feedback, resulting in significant fluctuations in mixing uniformity. Operators need to adjust mixing parameters based on experience, but the lack of precise monitoring methods makes it difficult to guarantee uniformity every time. While optimal mixing can be achieved, fluctuations in mixing uniformity can further affect the performance of the final eutectic structure, leading to unstable material properties. The technical solution provided in this application, through RFID batch identification linked to an intelligent metering feeder, acquires raw material purity and density data in real time. When raw material batch parameters deviate from preset values, the system automatically performs dynamic weight compensation and purity conversion to ensure that the actual effective components of the multi-component raw materials conform to the preset molar ratio, reducing the risk of component segregation caused by raw material fluctuations from the source. Simultaneously, a combination of high-energy ultrasound and electromagnetic mixing is employed, based on a deep reinforcement learning model, to analyze in real time the uniformity index obtained from in-situ LIBS and the temperature gradient data obtained from infrared thermal imaging and thermocouple networks. This dynamically adjusts mixing parameters such as ultrasonic frequency and electromagnetic field strength, forming a closed-loop control that effectively overcomes significant density differences between different components and improves mixing uniformity. Figure 1 As shown, it includes a feeding control module, a raw material mixing module, a mixing calculation module, and a reverse adjustment module.

[0047] Feeding control module: It acquires raw material weight data to obtain an initial weight set, identifies raw material batch information to obtain an initial information set, sets a central dynamic control system, acquires initial mixing information to obtain initial mixing items, and adds raw materials through the feeding component to obtain initial feeding items.

[0048] It should be noted that the initial information set corresponds to the initial weight set. A central dynamic control system is set up to obtain the initial mixing information, which includes the metering information of multiple component raw materials. The initial mixing item is obtained. Based on the initial mixing item, the raw materials are added through the feeding component to obtain the initial feeding item.

[0049] It is important to note that, such as Figure 2As shown, the method for obtaining the initial feed item includes: acquiring the weight data of multiple component raw materials in real time through a weight sensor to obtain an initial weight set, wherein the initial weight set includes at least one initial weight item consisting of the weight of one set of raw materials; identifying the batch information of the multiple component raw materials, including purity information and density information, using an RFID tag identification component to obtain an initial information set, wherein the initial information set includes at least one initial information item consisting of the purity information and density information of one set of raw materials; and based on the initial mixing item, acquiring the composition ratio information of the initial weight item and the initial information item through data acquisition, and then adding raw materials through a feeding component to obtain the initial feed item.

[0050] Specifically, such as Figure 3 As shown, the feeding component is an intelligent metering feeder, including a weight sensing unit and a batch identification unit, used to perform metering input of multi-component raw materials. The weight sensing unit consists of weight sensors, the number of which is the same as the number of portions of the multi-component raw materials. For example, when three types of multi-component raw materials are set, the weights of these three raw materials are obtained through three weight sensors respectively. The batch identification unit is an RFID tag identification device used to identify the batch information of the multi-component raw materials, including purity and density information. After obtaining the initial mixing information through the central dynamic control system, since the initial mixing information includes multi-component... The raw material metering information, such as the weight and purity of component 1 as a and aa, the weight and purity of component 2 as b and bb, and the weight and purity of component 3 as c and cc, allows for the addition of raw materials through an intelligent metering feeder, thus obtaining the initial feed items. The composition ratio information of the initial weight item and the initial information item is complementary. For example, for component 1, by adding weight and reducing purity, the weight and purity information of component 1 becomes a+ and aa-, and its composition ratio information is the same as a and aa, making the addition of raw materials more adjustable for preparation.

[0051] In the specific implementation process, the high-entropy eutectic ceramic material Al2O3-Y2O3-Ta2O5-ZrO2 is to be prepared. The initial mixing information is obtained by the central dynamic control system, as shown in Table 1.

[0052] Table 1

[0053]

[0054] At this point, an anomaly was detected in the Y2O3 batch via RFID batch identification. The measured purity was 99.92%, and the density was 5.03 g / cm³. Dynamic compensation calculations were then performed based on the initial weight and information ratios. A complementary balance was achieved through weight increases and purity conversion. The compensated weight = (target purity / actual purity) × target weight = (99.95% / 99.92%) × 183.5 ≈ 183.9 g. The purity conversion factor = actual purity / target purity = 99.92% / 99.95% ≈ 0.9997. Therefore, the raw material addition was adjusted, as shown in Table 2.

[0055] Table 2

[0056]

[0057] Based on the adjusted raw material addition data, the raw materials are added through an intelligent metering feeder to obtain the initial feed items.

[0058] Raw material mixing module: Set up a combined mixing device, and perform first and second mixing separately through the combined mixing device, and set the first initial value and the second initial value respectively to obtain the combined mixing item.

[0059] It should be noted that the combined mixing device includes a high-energy ultrasonic and electromagnetic mixing reaction component. The method for obtaining the combined mixing term includes: setting an initial frequency value and an initial field strength value, using the initial frequency value as the first initial value and the initial field strength value as the second initial value; performing ultrasonic oscillation with the initial frequency value and electromagnetic stirring with the initial field strength value, thereby obtaining the combined mixing term.

[0060] It should be noted that the methods for obtaining the initial frequency and initial field strength values ​​include: obtaining the working mixing degree of the high-energy ultrasound and electromagnetic mixing reaction components, including the ultrasonic oscillation mixing degree and the electromagnetic stirring mixing degree; based on the mapping relationship between the initial mixing term and the ultrasonic oscillation mixing degree and the electromagnetic stirring mixing degree, obtaining the oscillation mixing frequency set and the stirring mixing field strength set, wherein the oscillation mixing frequency set includes at least one oscillation mixing frequency term and the stirring mixing field strength set includes at least one stirring mixing field strength term, and the oscillation mixing frequency term and the stirring mixing field strength term correspond to each other to obtain at least one set of adjustment information; and using the oscillation mixing frequency term and the stirring mixing field strength term corresponding to the adjustment information set as the initial frequency value and the initial field strength value, respectively.

[0061] Specifically, such as Figure 4As shown, the working mixing degree of the high-energy ultrasonic and electromagnetic mixing reaction component is the ultrasonic oscillation mixing degree and the electromagnetic stirring mixing degree. The ultrasonic oscillation mixing degree and the electromagnetic stirring mixing degree achieve the target mixing state under different oscillation mixing frequencies and stirring mixing field strengths. For example, when it is necessary to mix multiple raw materials to a mixing homogeneity index ≥96%, the ultrasonic oscillation mixing degree at 80kHz exhibits a high-frequency cavitation effect, producing micron-sized bubble collapse and breaking up hard TaC aggregates. The electromagnetic stirring mixing degree at 0.8T forms a medium-strong eddy current field, resulting in laminar-turbulent flow. The measured mixing uniformity index was 96.5% when the mixture was in a transitional state, which suppressed the sinking of some raw materials. Alternatively, the ultrasonic oscillation mixing at 60 kHz allowed low-frequency, large-amplitude vibrations to penetrate deep deposits and loosen the interfacial bonds of some raw materials. The electromagnetic stirring mixing at 1.2 T generated Taylor vortices in a strong rotating magnetic field, which eliminated density stratification between some raw materials. The measured mixing uniformity index was 96.2%. Therefore, at least one of these groups was selected as the adjustment information group, and the oscillation mixing frequency term and stirring mixing field strength term corresponding to the adjustment information group were used as the initial frequency value and the initial field strength value, respectively.

[0062] It should be noted that the method for obtaining the initial frequency value and the initial field strength value also includes: obtaining the energy consumption data of the oscillation mixing frequency set and the stirring mixing field strength set respectively; sorting the energy consumption information corresponding to the adjustment information group in descending order to obtain the energy consumption ranking; obtaining the last information of the energy consumption ranking item as a control adjustment; and obtaining the oscillation mixing frequency item and the stirring mixing field strength item based on the control adjustment to obtain the initial frequency value and the initial field strength value.

[0063] Specifically, when the ultrasonic oscillation mixing degree and the electromagnetic stirring mixing degree are under different adjustment information groups, the energy consumption data are different. For example, when the energy consumption data of ultrasonic oscillation mixing degree a and electromagnetic stirring mixing degree b is 1, the energy consumption data of ultrasonic oscillation mixing degree c and electromagnetic stirring mixing degree d is 2, and the energy consumption data of ultrasonic oscillation mixing degree e and electromagnetic stirring mixing degree f is 3, and the ranking result of the energy consumption data is 1>2>3, the first information of the energy consumption ranking item is selected as the reference adjustment, that is, the ultrasonic oscillation mixing degree e and electromagnetic stirring mixing degree f are used as the initial frequency value and the initial field strength value.

[0064] Hybrid computing module: Based on the combined hybrid terms, it obtains hybrid real-time state information to obtain real-time information terms. Using real-time information terms as training data, it trains the model through the central dynamic control system, runs the deep reinforcement learning model, and outputs hybrid parameter adjustment instructions to obtain output adjustment terms.

[0065] It is important to note that the mixed real-time status information includes a uniformity index. The method for obtaining the real-time information item includes: setting up the acquisition device, which includes an in-situ laser-induced breakdown spectrometer; setting the pulsed laser index based on the in-situ laser-induced breakdown spectrometer, including the wavelength index and the pulse width index, to obtain the laser modulation item; setting the incident angle based on the laser modulation item; based on the incident angle, injecting the laser into the mixed powder layer to generate plasma by breaking down the surface; setting the capture range and at least two capture points; capturing the plasma emission spectrum through a spectral sensor to obtain the capture information item; and based on the capture information item, converting the spectral intensity into elemental concentration using the standard curve method and calculating the uniformity index to obtain the real-time information item.

[0066] Specifically, the pulsed laser index is set to 1064 nm, the wavelength index to 8 ns, and the laser modulation term is obtained. The incident angle is set to 90°, meaning it is perpendicularly incident on the mixed powder layer, breaking down the surface to generate plasma. Five capture points are set: one at the center and four at the edges. The spectral sensor captures the plasma emission spectrum in the range of 200-900 nm. The uniformity index is calculated using the following formula:

[0067] ;

[0068] in The uniformity index, , Let i be the concentration. This represents the average concentration.

[0069] It should be noted that the hybrid real-time status information also includes temperature gradient data. The method for obtaining the real-time information items includes: setting up an infrared thermal imaging array and an embedded thermocouple network to perform a three-dimensional temperature field scan, with at least two infrared thermal imaging arrays, scanning the surface temperature distribution of the combined hybrid device to obtain the scanned temperature item; based on the scanned temperature item, which is embedded in at least two embedded thermocouple networks inside the combined hybrid device, measuring the temperature of the core area to obtain the measured temperature item; and based on the position comparison information of the scanned temperature item and the measured temperature item, obtaining the temperature gradient map, and thus obtaining the real-time information item.

[0070] Specifically, there are six infrared thermal imaging arrays, evenly distributed on the surface of the combined hybrid device, and the embedded thermocouple network includes a thermocouple network consisting of 12 K-type armored thermocouples for measuring the temperature of the core area.

[0071] It is important to note that the method for obtaining the output adjustment term includes: using the initial feed item, the first initial value, the second initial value, and the real-time information item as input data, performing timestamp alignment and outlier cleaning to obtain the processing dataset; extracting core features based on the processing dataset, mapping the feature vectors to the state space of reinforcement learning, setting the algorithm framework and network structure, setting the reinforcement learning decision mechanism, including the definition of the action space and the design of the reward function, and finally generating and outputting parameter adjustment instructions to obtain the output adjustment term.

[0072] Specifically, timestamp alignment adopts IEEE The 1588 precision time protocol synchronizes data from multiple devices to microsecond-level errors. For example, the LIBS spectral timestamp is t=12:05:30.500, and power data is aligned to t=12:05:30.500±0.001s. Outlier cleaning is based on the 3σ principle to remove outliers, such as data with HUI>100% or <70%. Missing values ​​are filled by linear interpolation to compensate for transient signal loss. Core feature extraction includes multiple raw data and corresponding extracted features. For example, the extracted feature for elemental concentration is the concentration standard deviation, the extracted feature for power is the power change rate, and the extracted feature for HUI value is the HUI moving average. The algorithm framework is a near-end strategy optimization algorithm, and the network structure is an Actor-Critic dual network. In the defined dynamic space, different action commands represent different control parameters, such as a1 representing electromagnetic field strength, a2 representing ultrasonic frequency, etc. The parameter adjustment command generation and output are achieved through command generation logic, which scales and maps continuous action vectors to physical parameters, and determines the command output protocol through JSON structured data.

[0073] Reverse adjustment module: Based on the output adjustment term, the initial frequency value, initial field strength value and initial mixing term are adjusted, thereby realizing the preparation of multi-component high-entropy eutectic ceramic powder.

[0074] It is important to note that, in order to address the issues of uneven mixing and component segregation, a deep learning model is used to analyze sensor data in real time and dynamically adjust mixing parameters to achieve a self-optimizing mixing process. Closed-loop control is used to suppress segregation caused by density differences and ensure that the molar ratio error of high-entropy components is low.

[0075] A high-entropy eutectic ceramic powder material preparation device, using the aforementioned high-entropy eutectic ceramic powder material preparation control system, includes: a feeding component, comprising a weight sensing unit and a batch identification unit, for performing metering input of multi-component raw materials; a combined mixing device, for performing first and second mixing of the multi-component raw materials; a data acquisition unit, for acquiring real-time mixing status information of the multi-component raw materials; and a central dynamic control system, for running a deep reinforcement learning model and outputting mixing parameter adjustment instructions.

[0076] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended embodiments and their equivalents.

Claims

1. A control system for preparing high-entropy eutectic ceramic powder materials, comprising: Feeding control module: The feeding component executes the metering input of multi-component raw materials, obtains the raw material weight data to obtain the initial weight set, identifies the raw material batch information to obtain the initial information set, sets the central dynamic control system, obtains the initial mixing information through the central dynamic control system, the initial mixing information includes the metering information of multi-component raw materials to obtain the initial mixing item, and adds raw materials through the feeding component based on the initial mixing item to obtain the initial feeding item; Its characteristic is that it further includes: Raw material mixing module: Set up a combined mixing device, and perform first mixing and second mixing through the combined mixing device. Set a first initial value and a second initial value based on the first mixing and the second mixing, respectively. Perform first mixing with the first initial value and second mixing with the second initial value to obtain the combined mixing item; Hybrid computing module: Based on the combination of hybrid terms, it obtains real-time hybrid state information to obtain real-time information terms. Using real-time information terms as training data, it trains the model through the central dynamic control system, runs the deep reinforcement learning model, outputs hybrid parameter adjustment instructions to obtain output adjustment terms, and analyzes the training data in real time through the deep reinforcement learning model to dynamically adjust the first initial value and the second initial value, thereby realizing the self-optimization of the hybrid process. Reverse adjustment module: Based on the output adjustment term, it adjusts the initial frequency value, initial field strength value and initial mixing term. In order to address the problems of uneven mixing and component segregation, it uses a deep learning model to analyze sensor data in real time and dynamically adjust the mixing parameters. The central dynamic control system coordinates the entire process of feeding control module, raw material mixing module, mixing calculation module and reverse adjustment module to achieve multi-module integrated control. The method for obtaining the initial material input includes: The weight data of multiple raw materials are acquired in real time by a weight sensor to obtain an initial weight set, which includes at least one initial weight item consisting of the weight of a raw material. By combining RFID tag identification components, batch information of multi-component raw materials is identified, including purity information and density information, to obtain an initial information set. The initial information set includes at least one initial information item consisting of raw material purity information and density information. Based on the initial mixing item, the composition ratio information of the initial weight item and the initial information item is obtained through data acquisition, and then the raw materials are added through the feeding component to obtain the initial feeding item; The combined mixing device includes a high-energy ultrasonic and electromagnetic mixing reaction component, and the method for obtaining the combined mixing item includes: Set an initial frequency value and an initial field strength value, using the initial frequency value as the first initial value and the initial field strength value as the second initial value; Ultrasonic oscillation is performed at the initial frequency value, and electromagnetic stirring is performed at the initial field strength value, thereby obtaining a combined mixture term; The hybrid real-time state information includes a uniformity index, and the methods for obtaining the real-time information items include: The acquisition equipment is set up, including an in-situ laser-induced breakdown spectrometer. Based on the in-situ laser-induced breakdown spectrometer, the pulse laser index, including the wavelength index and the pulse width index, is set to obtain the laser adjustment term. The incident angle is set based on the laser adjustment term. Based on the incident angle, the laser is injected into the mixed powder layer to break through the surface and generate plasma. The capture range and at least two capture points are set. The plasma emission spectrum is captured by a spectral sensor to obtain the capture information. Based on the captured information item, the spectral intensity is converted into elemental concentration using the standard curve method, and the uniformity index is calculated to obtain the real-time information item; The hybrid real-time status information also includes temperature gradient data, and the methods for obtaining the real-time information items include: A three-dimensional temperature field scan is performed using an infrared thermal imaging array and an embedded thermocouple network. The number of infrared thermal imaging arrays is at least two. The surface temperature distribution of the combined hybrid device is scanned to obtain the scanned temperature item. Based on the scanned combined hybrid device with at least two embedded thermocouple networks embedded inside, the temperature of the core area is measured to obtain the measured temperature item. Based on the positional comparison information of the scanned temperature item and the measured temperature item, a temperature gradient map is obtained, and then real-time information items are obtained.

2. The high-entropy eutectic ceramic powder material preparation control system according to claim 1, wherein: The methods for obtaining the initial frequency value and the initial field strength value include: The working mixing degree of the high-energy ultrasound and electromagnetic mixing reaction components was obtained separately, including the ultrasonic oscillation mixing degree and the electromagnetic stirring mixing degree. Based on the mapping relationship between the initial mixing term and the ultrasonic oscillation mixing degree and the electromagnetic stirring mixing degree, the oscillation mixing frequency set and the stirring mixing field strength set are obtained. The oscillation mixing frequency set includes at least one oscillation mixing frequency term, and the stirring mixing field strength set includes at least one stirring mixing field strength term. The oscillation mixing frequency term and the stirring mixing field strength term correspond to each other, and at least one set of adjustment information is obtained. The oscillation mixing frequency term and the stirring mixing field strength term corresponding to the adjustment information group are used as the initial frequency value and the initial field strength value, respectively.

3. The system for controlling the preparation of a high-entropy eutectic ceramic powder material according to claim 2, characterized in that: The method for obtaining the initial frequency value and the initial field strength value also includes: Energy consumption data of the oscillation mixing frequency set and the stirring mixing field strength set are obtained respectively. The energy consumption information corresponding to the adjustment information group is sorted in descending order to obtain the energy consumption ranking. The last information of the energy consumption ranking item is obtained as a control adjustment. Based on the control adjustment, the oscillation mixing frequency item and the stirring mixing field strength item are obtained respectively, and then the initial frequency value and the initial field strength value are obtained.

4. The high-entropy eutectic ceramic powder material preparation control system according to claim 1, characterized in that: The method for obtaining the output adjustment term includes: Using the initial feeding item, the first initial value, the second initial value, and the real-time information item as input data, timestamp alignment and outlier cleaning are performed to obtain the processed dataset; Based on the processed dataset, core features are extracted, feature vectors are mapped to the state space of reinforcement learning, the algorithm framework and network structure are set, the reinforcement learning decision mechanism is set, including the definition of action space and the design of reward function, and finally the parameter adjustment instructions are generated and output, thus obtaining the output adjustment term.

5. A device for preparing high-entropy eutectic ceramic powder materials, characterized in that: The high-entropy eutectic ceramic powder material preparation control system according to any one of claims 1-4 includes: The feeding assembly includes a weight sensing unit and a batch identification unit, used to perform metering input of multi-component raw materials; A combined mixing device for first and second mixing of multi-component raw materials; The data acquisition unit is used to acquire real-time mixing status information of multiple component raw materials; The central dynamic control system is used to run deep reinforcement learning models and output mixed parameter adjustment commands.

Images