Integrated digital twinborn model construction method for ball mill

Abstract

The invention relates to a ball mill-oriented integrated digital twinborn model construction method, and belongs to the field of intelligence of mining mechanical equipment. The method comprises the following steps: firstly, constructing a digital twinborn model which comprises a dimension module, a scale module, a scene module and a service module; a digital twinborn model is constructed by taking a dimension module as a reference and embedding a scale module, a scene module and a service module. Thirdly, constructing a multi-sensor information acquisition system, and acquiring actual operation data of the ball mill; and finally, inputting the collected data and the constructed digital twinborn model into a consistency evaluation system, and performing consistency evaluation on the constructed integrated digital twinborn model. And a twinning model support is provided for monitoring the state of the ball mill based on digital twinning. And the current state of the physical entity ball mill is reflected in real time through the state change of the digital twin model in the operation process. And intelligent fault diagnosis of the ball mill is facilitated.

Description

Technical Field

[0001] The present invention relates to the field of intelligent mining machinery and equipment, and in particular to a method for constructing an integrated digital twin model for a ball mill. Background Art

[0002] Large-scale mining ball mills are the core equipment for mineral processing, and intelligent fault diagnosis is an inevitable trend in the development of the industry. Due to the complex structure, harsh operating environment, and huge workload of ball mills, potential safety hazards may occur during their operation. It is urgent to improve the intelligent status monitoring level of ball mills to achieve safe and reliable operation.

[0003] At present, the intelligent condition monitoring system based on sensors occupies the mainstream position. It monitors the operating status and fault characteristics of the ball mill through sensors. The intelligent condition monitoring system based on sensors can only meet the intelligent monitoring of the ball mill under specific requirements. This is because it can only rely on the collected historical data of the current equipment to judge the state of the ball mill. Since it is impossible to obtain sufficient fault data for the training of the diagnostic algorithm, the practical applicability of the designed monitoring system is low; and the intelligent condition monitoring system based on sensors cannot map the entire life cycle of the ball mill, which makes it impossible to consider the impact of the time variability of the ball mill on condition monitoring. Summary of the invention

[0004] The purpose of the present invention is to provide an integrated digital twin model construction method for ball mills, which solves the problems of low practical applicability and inability to fully map physical ball mills with high precision in existing monitoring systems. The present invention uses digital twin technology to assist in real-time status monitoring of ball mills throughout their life cycle, which can achieve the acquisition of typical fault data and full life cycle mapping of physical entity ball mills, and assist in real-time status monitoring of ball mills throughout their life cycle. The present invention involves scale modules, dimension modules, scenario modules, and service modules, forming an effective digital twin model construction system, which meets the needs of real-time status feedback of ball mills throughout their life cycle, and can provide a universal and effective modeling system for digital twin practitioners in various fields, promoting the comprehensive popularization and in-depth application of digital twins.

[0005] The above-mentioned object of the present invention is achieved by the following technical solutions: The method for constructing an integrated digital twin model for a ball mill includes the following steps: Step S1, constructing a ball mill digital twin model scale module; Step S2, constructing a ball mill digital twin model scenario module; Step S3, constructing a dimensional module of the ball mill digital twin model; Step S4, constructing a ball mill digital twin model service module; Step S5, constructing a ball mill multi-sensor information acquisition system; Step S6: construct a ball mill consistency evaluation system.

[0006] Furthermore, step S1 can be divided into spatial scale and disciplinary scale. According to the spatial scale, the construction of the digital twin model can be divided into three layers, namely, the part layer, the component layer and the equipment layer. In the process of constructing the digital twin model, the construction strategy from small to large and from local to global should be met. The part layer is constructed first, and then the component layer is constructed on the basis of the part layer, and finally the equipment layer is constructed; according to the disciplinary scale, the construction of the digital twin model can be divided into the unit layer, the component layer and the fusion layer. The unit layer is the most basic code input and basic operation. It is the most basic operation unit for constructing the digital twin model according to the disciplinary scale. Then, the component layer is obtained by combining the operation units according to different disciplines. Finally, multi-disciplinary fusion is considered to obtain the final fusion layer.

[0007] Furthermore, step S2 includes changes in the operating conditions of the ball mill and its own state; changes in the operating conditions of the ball mill involve changes in the operating parameters of the ball mill, including changes in rotation speed, changes in steel ball filling rate, changes in material filling rate, and changes in material type and particle size; the state of the ball mill itself involves its own faults and changes in its own structure.

[0008] Furthermore, step S3 includes models in four dimensions: geometry, physics, behavior, and rules, wherein: Construction of geometric model: The shape, size, structural composition and assembly relationship of the physical entity are fully reflected in the construction process. According to step S1, the construction of the geometric model can be divided into part-level geometric model, component-level geometric model and equipment-level geometric model; the part geometric model, as the bottom layer of the geometric model, is the basis for high-precision construction of the geometric model; the construction of the part geometric model mainly considers factors such as geometric size, geometric shape, geometric position and functional conditions; the construction of the component geometric model needs to be divided according to the different operating systems of the ball mill, and the parts of different systems are assembled into motor parts, transmission parts, cylinder parts, controller parts and sensor parts. In each assembly, the part-level geometric model is combined into a component-level geometric model according to the assembly relationship, assembly sequence and assembly position; the construction of the equipment-level geometric model needs to assemble the five component-level geometric models constructed above into an equipment-level geometric model based on the collaborative relationship and line connection; Construction of physical model: through the analysis of finite element, discrete element and multi-body dynamics, the operation process of the whole ball mill is simulated. According to step S1, the construction of physical model can be divided into unit-level physical model, component-level physical model and fusion-level physical model; unit-level physical model is the code input and basic operation in the analysis process; component-level physical model is the construction of finite element, discrete element or multi-body dynamics model, which is obtained on the basis of the construction of unit-level physical model; fusion-level physical model is the effective fusion of finite element model, discrete element model and multi-body dynamics model obtained in the construction of component-level physical model, and this process involves multidisciplinary coupling behavior; Construction of behavioral model: In-depth analysis of the basic characteristics of the ball mill, including mechanical structure, operating principle and key components. Based on step S2, the behavior of the ball mill in different scenarios is simulated according to actual work requirements, including but not limited to drum rotation behavior and grinding behavior; Construction of the rule model: Constraint rules, control rules and attenuation rules together constitute the rule model of the ball mill digital twin. The constraint rules require the actual working conditions of the ball mill to constrain the operating parameters, including speed constraints and filling rate constraints; the control rules need to apply appropriate control strategies based on the operating behavior of the ball mill; the attenuation rules are established using actual data collected from the physical ball mill in different time periods. This process requires the use of a deep neural network as an agent model.

[0009] Furthermore, step S4 requires injecting service functions into the established model. First, the communication between the physical ball mill and the digital twin model is established through the Internet of Things technology. Then, Unity 3D technology is used to build a modular and semantic front-end visualization interface with data visualization, status visualization and interactive and operable design functions.

[0010] Furthermore, step S5 includes a torque sensor, a vibration sensor, a noise sensor, a current sensor, a voltage sensor, an oil pressure sensor, a temperature sensor, a speed sensor, a stress measuring instrument, and a camera, which respectively measure the torque and speed of the ball mill main shaft, perform vibration detection on the ball mill transmission system, perform temperature measurement on the oil pressure and oil temperature of the lubricating oil, perform current and voltage measurement on the motor, and perform noise detection, stress measurement, and trajectory measurement on the ball mill cylinder.

[0011] Furthermore, step S6 includes consistency evaluation of geometry, physics, behavior, rules and integrated fusion models. The geometry model evaluation is to detect the accuracy of model establishment; the physical model evaluation is to evaluate the consistency of physical parameters of the ball mill, including current, voltage, vibration, noise, oil pressure, oil temperature and stress; the behavior model evaluation is to detect the consistency between the ball mill barrel operation trajectory and the digital twin model ball mill barrel operation trajectory; the rule model evaluation is to evaluate the consistency of parameter constraints and the consistency of ball mill speed under control rules; the integrated fusion model evaluation is to evaluate the consistency of power changes of the ball mill under different working conditions.

[0012] The beneficial effects of the present invention are: 1. Visual feedback on the real-time operating status of the ball mill throughout its life cycle: The Internet of Things technology is used to establish an interactive connection between the physical ball mill and the digital twin model. The digital twin model maps the physical entity with high precision and can reflect the current status of the physical ball mill in real time through the state changes of the digital twin model during operation.

[0013] 2. Interactive and operable functions: The digital twin model can be modified, adjusted and optimized through the interactive platform. The fault data information of the ball mill can be extracted by designing faults in the model to assist in the intelligent fault diagnosis of the ball mill. The model can also be adjusted to assist in the structural optimization of the model according to the state changes of the model.

[0014] 3. Comprehensive and high-precision model mapping: The dimension module is mainly based on the multidisciplinary modeling strategy of digital twins, that is, the integration of geometric models, physical models, behavioral models and rule models; the scale module refers to the model construction process that must meet the model construction strategy from small to large, from local to global, and from simple to complex; the scenario module refers to the construction of the model should consider the multi-condition operation characteristics of the ball mill body; the service module refers to the later stage of model construction, so that the constructed model has targeted service characteristics and gives the model specific service tasks. The fusion of multi-module information enables the digital twin model to map the physical entity ball mill with high precision. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The drawings described herein are used to provide further understanding of the present invention and constitute a part of this application. The illustrative examples of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0016] Figure 1 A system block diagram of the integrated digital twin model construction method for a ball mill according to the present invention; Figure 2 A diagram of a method for constructing an integrated digital twin model of a ball mill according to the present invention; Figure 3 A multi-sensor information acquisition system of the present invention; Figure 4 It is the consistency evaluation system of the present invention. DETAILED DESCRIPTION

[0017] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0018] See also Figures 1 to 4 As shown, the integrated digital twin model construction method for ball mill of the present invention first constructs a digital twin model, including a dimension module, a scale module, a scenario module and a service module. The digital twin model is constructed by embedding the dimension module, the scale module, the scenario module and the service module. Next, a multi-sensor information acquisition system is constructed to collect the actual operation data of the ball mill. Finally, the collected data and the constructed digital twin model are input into the consistency evaluation system, and the constructed integrated digital twin model is evaluated for consistency. Twin model support is provided for ball mill status monitoring based on digital twin.

[0019] See also Figure 1 As shown, the integrated digital twin model construction method for ball mill described in the present invention has the following specific implementation steps: Step S1, constructing a ball mill digital twin model scale module; Step S2, constructing a ball mill digital twin model scenario module; Step S3, constructing a dimensional module of the ball mill digital twin model; Step S4, constructing a ball mill digital twin model service module; Step S5, constructing a ball mill multi-sensor information acquisition system; Step S6: construct a ball mill consistency evaluation system.

[0020] See also Figure 2 As shown, the model construction of the present invention includes the following sub-modules: scale module, scenario module, dimension module, and service module; each sub-module is also composed of different components, the scale module is composed of spatial scale and subject scale, the scenario module is composed of changes in ball mill operating conditions and its own status, the dimension module is composed of models of four dimensions: geometry, physics, behavior, and rules, and the service module is composed of communication and visualization. Visualization can be further divided into data visualization, status visualization, and interactive operable design functions.

[0021] The scale module SAM is divided into a spatial scale SS and a subject scale DS, namely SAM= { SS , DS According to the spatial scale, the construction of the digital twin model can be divided into three layers, namely, the part layer PL, the component layer CL and the equipment layer EL, namely SS= { PL , CL , EL In the process of building a digital twin model, the construction strategy should be followed from small to large and from local to global. The part layer should be built first, and then the component layer should be built on the basis of the part layer, and finally the equipment layer should be built. According to the subject scale, the construction of the digital twin model can be divided into the unit layer UL, the component layer AL and the fusion layer FL, that is, DS ={ UL , AL , FL The unit layer is the most basic code input and basic operation. It is the most basic operation unit for building a digital twin model according to the subject scale. Then, the component layer is obtained by combining the operation units according to different disciplines. Finally, the final fusion layer is obtained by considering multidisciplinary integration.

[0022] The scenario module SEM includes the ball mill operating condition change COC and its own state SSC, that is, SEM ={ COC , SSC The change of ball mill operating conditions involves the change of ball mill operating parameters, including speed change RSC, steel ball filling rate change SRC, material filling rate change MRC, material category change MCC and material particle size change MSC, that is, COC ={ RSC , SRC , MRC , MCC , MSC The status of the ball mill itself involves its own fault SF and its own structural changes SC, that is, SSC ={ SF , SC}.

[0023] The dimensional module DMM includes four dimensional models: geometric model GM, physical model PM, behavioral model BM, and rule model RM, namely DMM ={ GM , PM , BM , RM},in: Construction of geometric model: The geometric model is constructed by using the 3D modeling software SolidWorks. The shape, size, structural composition and assembly relationship of the physical entity are fully reflected in the construction process. According to step S1, the construction of the geometric model can be divided into part-level geometric model PGM, component-level geometric model CGM and equipment-level geometric model EGM, that is, GM ={ PGM , CGM , EGM}; The part geometry model is the bottom layer of the geometry model and is the basis for high-precision construction of the geometry model; the main factors considered in the construction of the part geometry model include information such as geometry size GD, geometry shape GS, geometry position GP and functional condition FC, that is, PGM ={ GD , GS , GP , FC}, the part geometry model is divided into standard parts and custom parts during the construction process. Standard parts are imported from the standard parts library, and custom parts are generated according to specific features using modeling software. The features of the part geometry model (thread, material, etc.) are expressed based on size, shape, position and function to be closer to the actual situation of the physical entity; the construction of the component geometry model needs to be divided according to the different operating systems of the ball mill, and the parts of different systems are grouped into motor parts, transmission parts, barrel parts, controller parts and sensor parts. In each collection, the part-level geometry model is combined into a component-level geometry model according to the assembly relationship AR, assembly sequence AS and assembly position AP, that is, CGM ={ AR , AS , AP The construction of the device-level geometric model requires assembling the five component-level geometric models constructed above into a device-level geometric model based on the collaboration relationship CR and the line connection WC, that is, EGM ={ CR , WC}.

[0024] Construction of physical model: The operation process of the entire ball mill is simulated through the analysis of finite element, discrete element and multi-body dynamics. The finite element analysis uses Ansys software, the discrete element analysis uses EDEM software, and the multi-body dynamics analysis uses RecurDyn software. According to step S1, the construction of the physical model can be divided into unit-level physical model ULPM, component-level physical model CLPM and fusion-level physical model FLPM, that is, PM ={ ULPM , CLPM , FLPM}; The unit-level physical model is the code input and basic operations in the analysis process; the component-level physical model is the construction of the finite element, discrete element or multi-body dynamics model, which is obtained on the basis of the unit-level physical model construction; the fusion-level physical model is the effective fusion of the finite element model, discrete element model and multi-body dynamics model obtained in the component-level physical model construction. This process involves multidisciplinary coupling behavior.

[0025] Construction of behavioral model: In-depth analysis of the basic characteristics of the ball mill, including mechanical structure, operating principle and key components. Based on step S2, the ball mill behavior in different scenarios is simulated according to actual work requirements, including but not limited to cylinder rotation behavior and grinding behavior. Specific behavior settings can be set in EDEM and RecurDyn.

[0026] Construction of rule model: Constraint rules BR, control rules CR and attenuation rules AR together constitute the rule model of ball mill digital twin, that is, RM ={ BR , CR , AR}, the constraint rules require the actual working conditions of the ball mill to constrain the operating parameters, including speed constraints and filling rate constraints, which can be set in EDEM; the control rules can be implemented using Simulink, and appropriate control strategies need to be applied according to the operating behavior of the ball mill; the attenuation rules are established using the actual data collected from the physical ball mill in different time periods. By inputting data from different time periods into the residual network, the relationship between the data and the degree of attenuation can be established, and then the attenuation rules can be established.

[0027] The service module SRM first establishes communication between the physical ball mill and the digital twin model through the Internet of Things technology, using the TCP protocol as the communication protocol. Then, it uses Unity 3D to build a modular and semantic front-end visualization interface with data visualization, status visualization and interactive operable design functions.

[0028] See also Figure 3 As shown, the multi-sensor information acquisition system described in the present invention includes a torque sensor, a vibration sensor, a noise sensor, a current sensor, a voltage sensor, an oil pressure sensor, a temperature sensor, a speed sensor, a stress measuring instrument, and a camera. The torque sensor and the speed sensor are used to measure the torque and speed of the ball mill main shaft, the vibration sensor is used to detect the vibration of the ball mill transmission system, the oil pressure sensor and the temperature sensor are used to measure the oil pressure and oil temperature of the lubricating oil, the current sensor and the voltage sensor are used to measure the current and voltage of the motor, and the noise sensor, the stress measuring instrument and the camera are used to detect the noise, measure the stress and measure the trajectory of the ball mill cylinder.

[0029] See also Figure 4 As shown, the consistency evaluation system described in the present invention includes consistency evaluation of geometry, physics, behavior, rules and integrated fusion models. The geometry model evaluation is to detect the accuracy of model establishment; the physical model evaluation is to evaluate the consistency of physical parameters of the ball mill, including current, voltage, vibration, noise, oil pressure, oil temperature and stress; the behavior model evaluation is to detect the consistency between the running trajectory of the ball mill cylinder and the running trajectory of the ball mill cylinder of the digital twin model; the rule model evaluation is to evaluate the consistency of parameter constraints and the consistency of the ball mill speed under the control rules; the integrated fusion model evaluation is to evaluate the consistency of power changes of the ball mill under different working conditions.

[0030] The proposed ball mill digital twin modeling method and system were implemented on a ball mill prototype. The proposed method and system can establish a digital twin model of the ball mill with high precision. The accuracy of the established digital twin model was obtained by the consistency evaluation system and reached about 97%, which verified the effectiveness of the proposed digital twin modeling method and system.

[0031] The contents not described in detail in the specification of the present invention belong to the prior art known to the professional and technical personnel in this field.

[0032] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made to the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for constructing an integrated digital twin model for a ball mill, characterized in that: The following steps are involved: Step S1: constructing a ball mill digital twin model scale module; Step S2: constructing a ball mill digital twin model scenario module; Step S3: constructing a dimensional module of the ball mill digital twin model; Step S4: Construct the ball mill digital twin model service module.

2. The method for constructing an integrated digital twin model for a ball mill according to claim 1, characterized in that: The scale module for constructing the ball mill digital twin model described in step S1 is divided into spatial scale and disciplinary scale. According to the spatial scale, the construction of the digital twin model is divided into three layers, namely, the part layer, the component layer and the equipment layer. In the process of constructing the digital twin model, the construction strategy from small to large and from local to global is met. The part layer is constructed first, and then the component layer is constructed on the basis of the part layer, and finally the equipment layer is constructed; according to the disciplinary scale, the construction of the digital twin model is divided into a unit layer, a component layer and a fusion layer. The unit layer is the most basic code input and basic operation, and is the most basic operation unit for constructing the digital twin model according to the disciplinary scale. Then, the component layer is obtained by combining the operation units according to different disciplines, and finally, the multiple disciplines are integrated to obtain the final fusion layer.

3. The method for constructing an integrated digital twin model for a ball mill according to claim 1, characterized in that: The ball mill digital twin model scenario module constructed in step S2 includes changes in the operating conditions of the ball mill and its own status; changes in the operating conditions of the ball mill involve changes in the operating parameters of the ball mill, including changes in speed, changes in steel ball filling rate, changes in material filling rate, changes in material category and particle size; the ball mill's own status includes its own faults and changes in its own structure.

4. The method for constructing an integrated digital twin model for a ball mill according to claim 1, characterized in that: The dimension module for constructing the ball mill digital twin model in step S3 includes models of four dimensions: geometry, physics, behavior, and rules, wherein: Construction of geometric model: The shape, size, structural composition and assembly relationship of the physical entity are fully reflected in the construction process. According to step S1, the construction of the geometric model is divided into part-level geometric model, component-level geometric model and equipment-level geometric model; the part-level geometric model is the bottom layer of the geometric model and is the basis for high-precision construction of the geometric model; the factors considered in the construction of the part-level geometric model include geometric size, geometric shape, geometric position and functional condition information; the construction of the component-level geometric model needs to be divided according to the different operating systems of the ball mill, and the parts of different systems are assembled into motor parts, transmission parts, cylinder parts, controller parts and sensor parts. In each assembly, the part-level geometric models are combined into component-level geometric models according to the assembly relationship, assembly sequence and assembly position; the construction of the equipment-level geometric model is based on the collaborative relationship and line connection to assemble the five component-level geometric models constructed above into an equipment-level geometric model; Construction of physical model: through the analysis of finite element, discrete element and multi-body dynamics, the operation process of the whole ball mill is simulated, and the construction of the physical model is divided into unit-level physical model, component-level physical model and fusion-level physical model according to step S1; the unit-level physical model is the code input and basic operation in the analysis process; the component-level physical model is the construction of finite element, discrete element or multi-body dynamics model, which is obtained on the basis of the construction of unit-level physical model; the fusion-level physical model is the effective fusion of finite element model, discrete element model and multi-body dynamics model obtained in the construction of component-level physical model, and this process involves multidisciplinary coupling behavior; Construction of behavioral model: In-depth analysis of the basic characteristics of the ball mill, including mechanical structure, operating principle and key components. Based on step S2, the behavior of the ball mill in different scenarios is simulated according to actual work requirements, including drum rotation behavior and grinding behavior. Construction of the rule model: Constraint rules, control rules and attenuation rules together constitute the rule model of the ball mill digital twin. The constraint rules require the actual working conditions of the ball mill to constrain the operating parameters, including speed constraints and filling rate constraints; the control rules impose control strategies based on the operating behavior of the ball mill; the attenuation rules are established using actual data collected from the physical ball mill in different time periods, using a deep neural network as an agent model.

5. The method for constructing an integrated digital twin model for a ball mill according to claim 1, characterized in that: The ball mill digital twin model service module described in step S4 injects service functions into the established model. First, the communication between the physical ball mill and the digital twin model is established through the Internet of Things technology. Then, Unity 3D technology is used to build a modular and semantic front-end visualization interface with data visualization, status visualization and interactive and operable design functions.

Images