Hammer mill digital twin construction method based on multi-dimensional fusion modeling

Abstract

The application provides a hammer mill digital twin construction method based on multi-dimensional fusion modeling, comprising: constructing a multi-dimensional model of a hammer mill; the multi-dimensional model comprises: a geometric model, a physical model, a behavior model, a mechanism model and a data model; the multi-dimensional model is fused to construct a digital twin of the hammer mill. By constructing the geometric model, the physical model, the behavior model, the mechanism model and the data model, the structural properties of the hammer mill, the key physical quantities in the running process, the raw material particle crushing process, the mapping relationship between the key physical quantities and the equipment health index, and the change law of the crushing particle size in the production process are described, the performance and the crushing particle size of the hammer mill are obtained in real time and online, and important basis can be provided for keeping the hammer mill running smoothly and the stability of the crushing particle size.

Description

Technical Field

[0001] This invention relates to the field of manufacturing system automation technology, specifically to a method for constructing a digital twin of a hammer mill based on multi-dimensional fusion modeling, and a method for applying the digital twin of a hammer mill in production. Background Technology

[0002] Hammer mills are key equipment in the degerming and pulverizing section of fuel ethanol production lines, responsible for crushing crops such as corn, wheat, and cassava into granules. During production, the performance of the hammer mill is determined by multiple factors, including the performance of the variable frequency motor, the load on the main shaft, and the wear of the hammers. This performance not only affects the stability of the pulverized particle size, thus impacting subsequent processes and causing fluctuations in ethanol production quality, but also, excessive performance degradation can lead to equipment shutdown, reduced production efficiency, and greater economic losses for the company. Therefore, real-time monitoring of the hammer mill's performance and online acquisition of pulverized particle size are essential. However, currently, the performance of the hammer mill is estimated by on-site personnel based on the vibration amplitude and temperature of the variable frequency motor, and then judged based on experience. Particle size is also obtained through manual periodic sampling. Current methods are insufficient for accurately monitoring the hammer mill's performance in real time and for timely obtaining of pulverized particle size, resulting in instability in the production process. There is a need to develop a method for real-time online acquisition of the hammer mill's performance and pulverized particle size.

[0003] Digital twin technology can construct multi-dimensional fusion simulation models, analyze the correlations between historical data of equipment, and combine real-time sensor data to map the actual state of the equipment into a twin. Therefore, in recent years, this technology has received continuous attention from various industries and has been successfully implemented in fields such as aerospace, intelligent manufacturing, and smart cities. However, on the one hand, the application of digital twin technology is mostly concentrated on displaying synchronous virtual and real motion, with fewer applications in constructing equipment twins based on multi-dimensional fusion modeling; on the other hand, in process industries, especially in the fuel ethanol production industry, most equipment consists of variable frequency motors, pumps, and fans, and the actual state of the equipment often needs to be presented through multi-dimensional fusion modeling. Therefore, the application of digital twin technology in the fuel ethanol production industry is still relatively limited. Summary of the Invention

[0004] The purpose of this invention is to provide a method for constructing a digital twin of a hammer mill based on multi-dimensional fusion modeling. By constructing its geometric model, physical model, behavioral model, mechanistic model, and data model, the structural attributes of the hammer mill, key physical quantities during operation, raw material particle crushing process, mapping relationship between key physical quantities and equipment health index, and the variation law of crushed particle size during production are described. The performance and crushed particle size of the mill are obtained online in real time, which can provide an important basis for maintaining the stable operation of the mill and the stability of the crushed particle size.

[0005] To achieve the above objectives, the first aspect of the present invention provides a method for constructing a digital twin of a hammer mill based on multi-dimensional fusion modeling, the method comprising:

[0006] A multi-dimensional model of a hammer mill is constructed; the multi-dimensional model includes: a geometric model, a physical model, a behavioral model, a mechanistic model, and a data model;

[0007] By fusing the multi-dimensional models, a digital twin of the hammer mill is constructed.

[0008] Furthermore, the multi-dimensional models are fused to construct a digital twin of the hammer mill, including:

[0009] Load the geometric model as the basis for the digital twin;

[0010] The data acquisition module is designed to obtain key physical quantities of the hammer mill by accessing the real-time data interface provided by the physical model and associate them with the geometric model.

[0011] Design a behavior analysis module to integrate the behavior model into the digital twin using data analysis and visualization software;

[0012] The data analysis module is designed to use a visualization chart library to curve the mechanistic and data models and embed them into the digital twin, resulting in a digital twin of the hammer mill. The fused digital twin of the hammer mill can comprehensively perceive the mill's operating status, thereby guiding equipment maintenance strategies and production processes, ultimately achieving a closed loop of production, monitoring, optimization, and control.

[0013] Furthermore, the geometric model is established using the following method:

[0014] The structural properties of a hammer mill are defined formally as follows: GM_HM = {part_info, geom_info, asse_info, motion_info}

[0015] partInfo={base,motor,seat,rotor,hammer,...};

[0016] In the formula: GM_HM represents the geometric model of the hammer mill; part_info represents the component information of the mill, including at least the base, motor, seat, rotor, and hammer; geom_info represents the geometric dimension information; asse_info represents the assembly relationship of the mill; motion_info represents the relative motion of the mill.

[0017] A geometric model was created using 3D modeling software based on the structural properties described. The structural properties of the geometric model were obtained by surveying and mapping each component of the hammer mill. The assembly relationships and relative movements of each component are the same as those of the hammer mill, and the constructed geometric model can fully represent the hammer mill.

[0018] Furthermore, the method for establishing the geometric model also includes:

[0019] The geometric model is subjected to lightweighting and rendering to obtain a geometric model with the same appearance as the hammer mill. Lightweighting retains only the main components and essential features of the model, greatly reducing its size while maintaining fidelity. Rendering the model and applying textures to its surface results in an appearance that is essentially identical to the actual equipment, achieving a visual twin.

[0020] Furthermore, the physical model is established using the following method:

[0021] Multiple sensors are arranged on the hammer mill to monitor key physical quantities of the hammer mill. These key physical quantities are represented as follows:

[0022] PSet = {I,n,V,T};

[0023] In the formula: PSet is the real-time data set of key physical quantities; I is the variable frequency motor current; n is the main shaft speed; V is the vibration dataset, which includes real-time vibration data from 6 measuring points of the motor and bearings; T is the temperature dataset, which includes real-time temperature data from 6 measuring points of the motor and bearings. The constructed physical model can display the changes of key physical quantities such as I, n, V, and T in real time during the operation of the crusher, which reflects the performance of the hammer mill.

[0024] Furthermore, the behavioral model includes the motion law of the raw material in the grinding chamber and the deformation law of the raw material in the grinding chamber. The behavioral model is established by the following method:

[0025] Historical data on particle size under different influencing factors were analyzed, and each factor was normalized.

[0026] After normalization, the factors are ranked by importance, and the top six factors are taken as the factors affecting particle size, as follows:

[0027] <speed,type rm Moisture rm ,size,wear,current>;

[0028] Where: speed is the feed rate; type rm For raw material types; moisture rm The moisture content of the raw material is represented by the sieve aperture size, wear by the hammer wear, and current by the main shaft current.

[0029] The factors affecting the particle size of the pulverized material were simulated and analyzed using multibody dynamics simulation software, and the motion law of the raw material in the pulverizing chamber was obtained.

[0030] The factors affecting particle size were simulated and analyzed using finite element analysis software to obtain the deformation law of the raw material in the grinding chamber. Based on the factors affecting particle size obtained from historical data, the motion and deformation laws of the raw material in the grinding chamber were simulated and analyzed based on these factors. The material grinding process and particle size can be reflected in real time in the digital twin of the hammer mill.

[0031] Furthermore, the mechanistic model is established through the following method:

[0032] Define the equipment health index H, with a value range of [0, 100];

[0033] Analyze historical data of key physical quantities and the status of the hammer mill;

[0034] A nonlinear mapping relationship between key physical quantities and equipment health indices is constructed using the support vector machine method:

[0035] H = F(I,n,V,T);

[0036] 0≤H≤100;

[0037] A mechanistic model is constructed based on the nonlinear mapping relationship between key physical quantities and equipment health indices. This model can effectively predict and assess the health status of hammer mills, monitor equipment performance in real time, and formulate different maintenance strategies.

[0038] Furthermore, the data model is established using the following method:

[0039] A LSTM-based particle size prediction model was constructed as the data model based on historical particle size data. Statistical analysis of the historical particle size data revealed the variation pattern of particle size during production. This variation pattern implies the time-varying nature of equipment performance. The LSTM-based particle size prediction model can predict the particle size for subsequent time steps based on previous particle size data, thereby enabling timely production adjustments.

[0040] Furthermore, the method also includes:

[0041] The following quadruple is defined for the scientific evaluation of digital twins:

[0042] <accuracy,confidence,delay,stability> ;

[0043] Where: accuracy is the modeling accuracy of the digital twin; confidence is the simulation confidence of the digital twin; delay is the virtual-real synchronization delay; and stability is the operational stability of the digital twin.

[0044] A second aspect of this invention provides a method for applying a digital twin of a hammer mill in production, wherein the digital twin of the hammer mill is constructed using a hammer mill digital twin construction method based on multi-dimensional fusion modeling; the application method includes:

[0045] The digital twin of the hammer mill is deployed on the server of the production control center;

[0046] Key physical quantities are obtained in real time through the physical model in the digital twin of the hammer mill.

[0047] The motion and deformation patterns of the raw materials in the crushing chamber are obtained in real time through the behavior model in the digital twin of the hammer mill based on key physical quantities.

[0048] The performance of the hammer mill is predicted in real time based on key physical quantities using the mechanism model in the digital twin of the hammer mill.

[0049] The data model in the digital twin of the hammer mill predicts the particle size of the material in real time based on the deformation pattern of the raw material in the crushing chamber. Combined with historical data in the database, equipment performance / particle size can be predicted, thereby optimizing the production process of the crushing section. Based on the above optimization results, the actual production process is controlled, ultimately completing a closed loop of production, monitoring, optimization, and control.

[0050] Through the above technical solutions, a digital twin of the hammer mill is constructed from the perspective of multi-dimensional fusion modeling. By analyzing the correlation between historical data of the equipment and combining real-time sensor data, the actual status of the equipment is mapped by the twin. The performance and particle size of the hammer mill are obtained online in real time, thereby guiding the equipment maintenance strategy and production process, and finally realizing a closed loop of production, monitoring, optimization and control.

[0051] Other features and advantages of the embodiments of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0052] The accompanying drawings are provided to further illustrate embodiments of the present invention and form part of the specification. They are used together with the following detailed description to explain the embodiments of the present invention, but do not constitute a limitation thereof. In the drawings:

[0053] Figure 1 This is a simplified assembly diagram of a hammer mill.

[0054] Figure 2 This is a working principle diagram of a hammer mill.

[0055] Figure 3 This is a flowchart of a method for constructing a digital twin of a hammer mill based on multi-dimensional fusion modeling, provided by one embodiment of the present invention.

[0056] Figure 4 It is a geometric model diagram that has undergone lightweighting and rendering processes;

[0057] Figure 5 This is a schematic diagram of the mechanism of a hammer mill.

[0058] Figure 6 This is a data model diagram of a hammer mill;

[0059] Figure 7 This is a flowchart illustrating the process of constructing a digital twin of a hammer mill by fusing multi-dimensional models in a method for constructing a digital twin of a hammer mill based on multi-dimensional fusion modeling, provided by one embodiment of the present invention.

[0060] Figure 8 A schematic diagram of the operation of a digital twin of a hammer mill, established according to one embodiment of the present invention.

[0061] Explanation of reference numerals in the attached figures

[0062] 1-Feed guide plate, 2-Wall plate, 3-Reinforcing ring, 4-Rotor, 5-Hammer blade, 6-Operating door, 7-Guard frame, 8-Bearing seat, 9-Base. Detailed Implementation

[0063] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0064] This application focuses on hammer mills as the research object, such as Figure 1 As shown, the hammer mill includes components such as a feed guide plate 1, wall plate 2, reinforcing ring 3, rotor 4, hammers 5, operating door 6, protective frame 7, bearing seat 8, and base 9. The working principle of the hammer mill is as follows: Figure 1 As shown, the material to be crushed is fed in through the top inlet and enters the crushing chamber from the left or right through the feed guide plate. Under the impact of the high-speed rotating hammer and the friction of the screen plate, the material is gradually crushed and discharged from the bottom outlet through the screen holes under the action of centrifugal force and airflow.

[0065] Figure 3 This is a flowchart of a method for constructing a digital twin of a hammer mill based on multi-dimensional fusion modeling, according to one embodiment of the present invention. Figure 3 As shown, the method includes:

[0066] Step 1: Construct a multi-dimensional model of the hammer mill; the multi-dimensional model includes: geometric model, physical model, behavioral model, mechanistic model and data model.

[0067] In this embodiment, the geometric model is established using the following method:

[0068] The structural properties of the hammer mill are defined formally as follows: GM_HM = {part_info, geom_info, asse_info, motion_info};

[0069] partInfo={base,motor,seat,rotor,hammer,...};

[0070] In the formula: GM_HM represents the geometric model of the hammer mill; part_info represents the component information of the mill, including the base, variable frequency motor, seat, rotor, hammer, etc.; geom_info represents the geometric dimension information; asse_info represents the assembly relationship of the mill; motion_info represents the relative motion of the mill.

[0071] A geometric model is created using 3D modeling software based on the structural properties. In some embodiments, the geometric model is created using CAD modeling software such as SolidWorks or UG. The structural properties of the geometric model are obtained by surveying and mapping each component of the hammer mill. The assembly relationships and relative movements of each component are the same as those of the hammer mill. The constructed geometric model can fully represent the hammer mill and reflect its structural properties, serving as the basis for the construction of a digital twin.

[0072] The geometric models built by CAD modeling software are often quite large, and their surface textures do not match the actual equipment. In some embodiments, the geometric models built by the modeling software are further lightweighted and rendered to obtain a geometric model with the same appearance as a hammer mill. Lightweighting retains only the main components and essential features of the model, significantly reducing its size while maintaining fidelity. Model rendering applies textures to the surface of the geometric model, resulting in an appearance that is essentially identical to the actual equipment, achieving a visual twin. For example... Figure 4 As shown, the geometric model constructed by the CAD modeling software has been lightweighted and rendered, reducing the model size from 74MB to 1.45MB. At the same time, the outer tube effect is more realistic, thus greatly reducing the model size while maintaining fidelity.

[0073] In this embodiment, the physical model is established using the following method:

[0074] Multiple sensors are arranged on the hammer mill to monitor key physical quantities of the hammer mill. These key physical quantities are represented as follows:

[0075] PSet = {I,n,V,T};

[0076] In the formula: PSet is the real-time data set of key physical quantities; I is the variable frequency motor current; n is the spindle speed; V is the vibration dataset, containing real-time vibration data from 6 measuring points of the motor and bearings; T is the temperature dataset, containing real-time temperature data from 6 measuring points of the motor and bearings. Figure 6 As shown, the constructed physical model can display the changes in key physical quantities such as I, n, V, and T in real time during the operation of the crusher, which reflects the performance of the hammer crusher.

[0077] In this embodiment, the behavioral model includes the motion law of the raw material in the grinding chamber and the deformation law of the raw material in the grinding chamber. The behavioral model is established by the following method:

[0078] Historical data on particle size under different influencing factors were analyzed, and each factor was normalized.

[0079] After normalization, the factors are ranked by importance, and the top six factors are taken as the factors affecting particle size, as follows:

[0080] <speed,type rm Moisture rm ,size,wear,current>;

[0081] Where: speed is the feed rate; type rm For raw material types; moisture rm The moisture content of the raw material is represented by the sieve aperture size, wear by the hammer wear, and current by the main shaft current.

[0082] The factors affecting the particle size of the pulverized material were simulated and analyzed using the multibody dynamics simulation software ADAMS, and the motion law of the raw material in the pulverizing chamber was obtained.

[0083] The factors affecting particle size were simulated and analyzed using the finite element analysis software ABAQUS to obtain the deformation law of the raw material in the grinding chamber. Based on the factors affecting particle size obtained from historical data, the motion and deformation laws of the raw material in the grinding chamber were simulated and analyzed based on these factors, reflecting the production behavior of the hammer mill. The material grinding process and particle size can be reflected in real time in the digital twin of the hammer mill.

[0084] In this embodiment, as Figure 5 As shown, the mechanism model is established through the following method:

[0085] The equipment health index H is defined, with a value range of [0, 100]. The closer the value of H is to 100, the better the health of the crusher; conversely, the closer the value of H is to 0, the worse the health of the crusher.

[0086] Analyze historical data of key physical quantities and the status of the hammer mill;

[0087] A nonlinear mapping relationship between key physical quantities and equipment health indices is constructed using the Support Vector Machine (SVM) method.

[0088] H = F(I,n,V,T);

[0089] 0≤H≤100;

[0090] A mechanistic model is constructed based on the nonlinear mapping relationship between key physical quantities and equipment health indices. This model can effectively predict and assess the health status of hammer mills, monitor equipment performance in real time, and formulate different maintenance strategies.

[0091] In this embodiment, the data model is established using the following method:

[0092] A particle size prediction model based on a Long Short-Term Memory (LSTM) network is constructed as the data model based on historical particle size data. Leveraging the suitability of LSTM for time series forecasting, statistical analysis of historical particle size data reveals the variation patterns in particle size during production. These patterns implicitly represent the time-varying nature of equipment performance. The LSTM-based particle size prediction model can predict the particle size for subsequent time steps based on previous data, enabling timely production adjustments.

[0093] In one embodiment, the raw material is corn, such as... Figure 6 As shown, 900 historical data points of crushed particle size from the past 5 months were selected as training samples for a Long Short-Term Memory (LSTM) network for statistical analysis. The time step of this time series was 4 hours, and the variation pattern of crushed particle size during this period was obtained.

[0094] It should be noted that hammer mills can process raw materials including crops such as corn, wheat, and cassava. The factors affecting the crushing force of different raw materials are not the same and need to be analyzed independently. The historical data used should also be historical data of the same raw material production process.

[0095] Step Two: Merge the multi-dimensional models to construct a digital twin of the hammer mill (HMDT), such as... Figure 7 As shown, it specifically includes:

[0096] S201: Load the geometric model as the basis of the digital twin. In this embodiment, the various components of the geometric model are obtained in the babylon.js web-based visualization framework, serving as the basis of the digital twin;

[0097] S202: Design a data acquisition module to obtain key physical quantities of the hammer mill by accessing the real-time data interface provided by the physical model and associate them with the geometric model. In this embodiment, the data acquisition module is built in the babylon.js web-based visualization framework to access the real-time data interface provided by the physical model, obtain key physical quantities of the hammer mill, and achieve virtual-real synchronization between the hammer mill and its digital twin.

[0098] S203: Design a behavior analysis module to integrate the behavior model into the digital twin using data analysis and visualization software. In this embodiment, the behavior model is integrated into the digital twin using ParaView data analysis and visualization software, providing the digital twin with simulation capabilities and enabling the analysis of the impact of different physical quantities in the physical model on the raw material crushing process.

[0099] S204: Design a data analysis module, using a visualization chart library to curve the mechanism model and data model and embed them into the digital twin, thus obtaining a digital twin of the hammer mill. For example... Figure 8 The diagram shows a schematic of the digital twin constructed by the method of this application. In the digital twin constructed by this invention, the mechanistic model is represented by a device performance curve. This curve is drawn using the echarts.js visualization chart library and embedded into the digital twin, characterizing the variation of device performance with key physical quantities in the physical model. The data model is represented by a particle size statistical curve, which is also drawn using the echarts.js visualization chart library and embedded into the digital twin, allowing prediction of the particle size at the next moment based on historical statistical data. The hammer mill digital twin formed by the above steps can comprehensively perceive the working status of the mill, thereby guiding equipment maintenance strategies and production processes, ultimately achieving a closed loop of production, monitoring, optimization, and control.

[0100] It should be noted that, in this application, if Figure 8 As shown, the real-time and historical data of physical quantities, as well as the historical and real-time data of different material particle sizes, are all stored in the equipment database and can be retrieved directly from the equipment database when in use.

[0101] In some other embodiments, the initially constructed digital twin still needs to undergo a scientific evaluation of its practicality. In this embodiment, the following quadruple is defined for the scientific evaluation of the digital twin:

[0102] <accuracy,confidence,delay,stability> ;

[0103] Where: accuracy is the modeling accuracy of the digital twin; confidence is the simulation confidence of the digital twin; delay is the virtual-real synchronization delay; and stability is the operational stability of the digital twin.

[0104] The digital twin of the hammer mill constructed using the above method is expected to achieve the following performance indicators after deployment:

[0105] <90%, 85%, 100ms, 95%>.

[0106] Based on the above assessment results, it can be concluded that HMDT can meet the digital twin requirements of hammer mills in production sites.

[0107] A second aspect of this invention provides a method for applying a digital twin of a hammer mill in production, wherein the digital twin of the hammer mill is constructed using a hammer mill digital twin construction method based on multi-dimensional fusion modeling; the application method includes:

[0108] The digital twin of the hammer mill is deployed on the server of the production control center;

[0109] Key physical quantities are obtained in real time through the physical model in the digital twin of the hammer mill.

[0110] The motion and deformation patterns of the raw materials in the crushing chamber are obtained in real time through the behavior model in the digital twin of the hammer mill based on key physical quantities.

[0111] The performance of the hammer mill is predicted in real time based on key physical quantities using the mechanism model in the digital twin of the hammer mill.

[0112] The data model in the digital twin of the hammer mill predicts the particle size of the material in real time based on the deformation pattern of the raw material in the crushing chamber. Combined with historical data in the database, equipment performance / particle size can be predicted, thereby optimizing the production process of the crushing section. Based on the above optimization results, the actual production process is controlled, ultimately completing a closed loop of production, monitoring, optimization, and control.

[0113] This invention relates to a method for constructing a digital twin of a hammer mill based on multi-dimensional fusion modeling. It constructs a digital twin of the hammer mill from the perspective of multi-dimensional fusion modeling, and by analyzing the correlation between historical data of the equipment and combining real-time sensor data, the actual state of the equipment is mapped to the twin. The performance and particle size of the hammer mill are obtained online in real time, thereby guiding equipment maintenance strategies and production processes, and ultimately realizing a closed loop of production, monitoring, optimization and control.

[0114] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. This program is stored in a storage medium and includes several instructions to cause a microcontroller, chip, or processor to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0115] The optional embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the embodiments of the present invention are not limited to the specific details described above. Within the scope of the technical concept of the embodiments of the present invention, various simple modifications can be made to the technical solutions of the embodiments of the present invention, and these simple modifications all fall within the protection scope of the embodiments of the present invention. It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, the embodiments of the present invention will not further describe the various possible combinations.

[0116] Furthermore, various different embodiments of the present invention can be combined in any way, as long as they do not violate the spirit of the embodiments of the present invention, they should also be regarded as the content disclosed by the embodiments of the present invention.

Claims

1. A method for constructing a digital twin of a hammer mill based on multi-dimensional fusion modeling, characterized in that, The method includes: A multi-dimensional model of a hammer mill is constructed; the multi-dimensional model includes: a geometric model, a physical model, a behavioral model, a mechanistic model, and a data model; By fusing the multi-dimensional models, a digital twin of the hammer mill is constructed. The behavioral model includes the motion law of the raw material in the grinding chamber and the deformation law of the raw material in the grinding chamber. The behavioral model is established by the following method: Historical data on particle size under different influencing factors were analyzed, and each factor was normalized. After normalization, the factors were ranked by importance, and the top six factors were selected as the factors affecting the particle size. The factors affecting the particle size of the pulverized material were simulated and analyzed using multibody dynamics simulation software, and the motion law of the raw material in the pulverizing chamber was obtained. The factors affecting the particle size of the pulverized material were simulated and analyzed using finite element analysis software, and the deformation law of the raw material in the pulverizing chamber was obtained.

2. The method for constructing a digital twin of a hammer mill based on multi-dimensional fusion modeling according to claim 1, characterized in that, The multi-dimensional models are fused to construct a digital twin of the hammer mill, including: Load the geometric model as the basis for the digital twin; The data acquisition module is designed to obtain key physical quantities of the hammer mill by accessing the real-time data interface provided by the physical model and associate them with the geometric model. Design a behavior analysis module to integrate the behavior model into the digital twin using data analysis and visualization software; The data analysis module is designed, and a visualization chart library is used to curve the mechanism model and data model and embed them into the digital twin to obtain the digital twin of the hammer mill.

3. The method for constructing a digital twin of a hammer mill based on multi-dimensional fusion modeling according to claim 1, characterized in that, The geometric model is established using the following method: The structural properties of a hammer mill are defined formally as follows: ; ； In the formula: A geometric model representing a hammer mill; Information on the components of the shredder, including at least the base. Variable frequency motor Bearing housing Crusher rotor Hammer blade part; Represents geometric dimension information; This indicates the assembly relationship of the crusher; Indicates the relative motion of the crusher; A geometric model is created using 3D modeling software based on the structural properties.

4. The method for constructing a digital twin of a hammer mill based on multi-dimensional fusion modeling according to claim 3, characterized in that, The method for establishing the geometric model also includes: The geometric model is then subjected to lightweighting and rendering processes to obtain a geometric model with the same appearance as a hammer mill.

5. The method for constructing a digital twin of a hammer mill based on multi-dimensional fusion modeling according to claim 1, characterized in that, The physical model was established using the following method: Multiple sensors are arranged on the hammer mill to monitor key physical quantities of the hammer mill, which are expressed as follows: ； In the formula: A collection of real-time data for key physical quantities; This refers to the current of the variable frequency motor. Main spindle speed; This is a vibration dataset, containing real-time vibration data from six measuring points on the motor and bearing. This is a temperature dataset, containing real-time temperature data from six measuring points on the motor and bearing.

6. The method for constructing a digital twin of a hammer mill based on multi-dimensional fusion modeling according to claim 1, characterized in that, The factors affecting the particle size of the pulverized material are expressed as follows: ； in: For feeding speed; Types of raw materials; For the moisture content of raw materials, For sieve aperture size, For hammer wear, It is the main spindle current.

7. The method for constructing a digital twin of a hammer mill based on multi-dimensional fusion modeling according to claim 5, characterized in that, The mechanistic model was established using the following method: Define the equipment health index H, with a value range of 100%. ; Analyze historical data of key physical quantities and the status of the hammer mill; A nonlinear mapping relationship between key physical quantities and equipment health indices is constructed using the support vector machine method: ； ； A mechanism model is constructed based on the nonlinear mapping relationship between key physical quantities and equipment health index.

8. The method for constructing a digital twin of a hammer mill based on multi-dimensional fusion modeling according to claim 1, characterized in that, The data model was established using the following method: A LSTM-based particle size prediction model was constructed as a data model based on historical particle size data.

9. The method for constructing a digital twin of a hammer mill based on multi-dimensional fusion modeling according to claim 1, characterized in that, The method further includes: The following quadruple is defined for the scientific evaluation of digital twins: ； in: To improve the modeling accuracy of digital twins; The confidence level of the digital twin simulation; This is to delay the synchronization between virtual and real systems; To ensure the operational stability of the digital twin.

10. A method for applying a digital twin of a hammer mill in production, characterized in that, The digital twin of the hammer mill is constructed using the hammer mill digital twin construction method based on multi-dimensional fusion modeling as described in any one of claims 1-9; the application method includes: The digital twin of the hammer mill is deployed on the server of the production control center; Key physical quantities are obtained in real time through the physical model in the digital twin of the hammer mill. The motion and deformation patterns of the raw materials in the crushing chamber are obtained in real time through the behavior model in the digital twin of the hammer mill based on key physical quantities. The performance of the hammer mill is predicted in real time based on key physical quantities using the mechanism model in the digital twin of the hammer mill. The data model in the digital twin of the hammer mill is used to predict the particle size of the material in real time based on the deformation law of the raw material in the crushing chamber.

Images