Mine semi-autogenous grinding process analog simulation method and system based on discrete elements

Abstract

The invention discloses a mine semi-autogenous grinding process analog simulation method and system based on discrete elements, and relates to the field of analog simulation, and the method comprises the steps: according to a crushing characteristic label and a parameter mapping matrix, fusing Internet of Things sensor data, constructing a particle size distribution prediction model and an energy consumption optimization scheduling strategy, and obtaining an energy-saving operation plan; comparing the simulation data set with field production data, adjusting multi-phase coupling particle model parameters, generating a calibration parameter set, researching the influence law of semi-autogenous mill operation parameters by adopting a control variable method based on the calibration parameter set, obtaining a performance index database and an analysis report, and fusing multi-sensor data and laser radar data to obtain a semi-autogenous mill operation parameter analysis result. An ore grinding process sensing mechanism and optimization control logic are constructed, and a visual interface is generated; through multiphase modeling and data driving optimization, the simulation precision and the process suitability are comprehensively improved, and an efficient and economical solution is provided for the mine semi-autogenous grinding process.

Description

Technical Field

[0001] The present invention relates to the field of simulation, and in particular to a method and system for simulating a mining semi-autogenous grinding process based on discrete elements. Background Art

[0002] As a core technology in the field of mineral processing, the semi-autogenous grinding process in mines is widely used in the crushing and grinding of metal ores and non-metallic ores. Its efficiency directly affects the energy consumption and output of the mineral processing process. In recent years, with the improvement of computing power and the development of numerical simulation technology, the discrete element method has become an important tool for simulating the movement and crushing behavior of particles in semi-autogenous grinding mills. Traditional discrete element simulation establishes a particle geometry model and combines it with a contact mechanics model to simulate the collision and energy transfer between particles, generate motion trajectories and crushing data, and provide support for optimizing mill operating parameters. At the same time, industrial CT scanning and three-dimensional reconstruction technology are gradually applied to ore microstructure analysis. By obtaining the fracture network and mineral distribution characteristics, they provide a geometric basis for particle models. These technological advances have promoted the digital transformation of the semi-autogenous grinding process and laid the foundation for improving fine particle yield, reducing energy consumption and extending equipment life.

[0003] However, the simulation accuracy and optimization capabilities of existing technologies are still limited by the assumption of a single particle model, static operating parameter settings, and limited real-time monitoring methods, making it difficult to fully adapt to complex ore types and dynamic operating conditions. First, traditional discrete element simulations generally use a uniform particle model, ignoring the heterogeneity of the ore's internal microscopic fracture network, bedding distribution, and mineral embedding characteristics. This leads to a large deviation between the simulation results and the actual crushing behavior, making it difficult to accurately predict the crushing characteristics of different mineral regions. In addition, existing methods mostly rely on static empirical formulas or single operating condition analysis to optimize operating parameters, without comprehensive consideration of dynamic adaptability to multiple operating conditions and energy consumption optimization, which limits the effectiveness of increasing fine particle yield and reducing energy consumption. Summary of the Invention

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

[0005] Therefore, the present invention provides a method and system for simulating a mining semi-autogenous grinding process based on discrete elements to solve the problem of limiting the improvement of fine particle yield and the reduction of energy consumption.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: In a first aspect, the present invention provides a method for simulating a mining semi-autogenous grinding process based on discrete elements, which includes: The microscopic fracture network, bedding distribution, and mineral embedding characteristics of the ore samples are collected. A digital twin library of the ore is generated through a 3D reconstruction algorithm. The prototypes of the irregular particles are then segmented and assigned non-uniform physical property parameters to construct a multi-phase coupled particle model. Based on a multiphase coupled particle model, a dynamic variable boundary surface is updated in a discrete element simulation framework to track the particle motion trajectories and collision patterns within the semi-autogenous mill. After generating a simulation dataset, a deep learning algorithm is used to mine the inter-particle contact force network topology and energy transfer paths, obtain the crushing characteristic labels, and generate a parameter mapping matrix. Based on the crushing characteristic labels and parameter mapping matrix, IoT sensor data is integrated to build a particle size distribution prediction model and energy consumption optimization scheduling strategy to obtain an energy-saving operation plan. The simulation data set is compared with the field production data, and the parameters of the multiphase coupled particle model are adjusted to generate a calibration parameter set. Based on the calibration parameter set, the control variable method is used to study the influence of the semi-autogenous grinding mill operating parameters, obtain a performance indicator database and analysis report, and integrate multi-sensor data and lidar data to build a grinding process perception mechanism and optimize control logic, generating a visual interface. Based on the visual interface, the optimized control logic is verified on site, the multiphase coupled particle model and optimized control logic are iteratively updated, and an implementation report is generated.

[0007] As a preferred solution of the discrete element-based simulation method for semi-autogenous grinding in mines described in the present invention, the microscopic fracture network, bedding distribution, and mineral embedding characteristics of the ore sample are collected, and the ore digital twin library is generated through a three-dimensional reconstruction algorithm. After that, the prototype of the irregular particles is segmented and assigned with non-uniform physical property parameters to construct a multiphase coupled particle model, specifically: Obtain representative ore samples from the mine site, perform 3D scanning, and collect projection image data including microscopic cracks, bedding structures, and mineral phase boundaries; Import the projection image data into the 3D reconstruction software to generate 3D volume data, and then form a 3D digital model through preprocessing; Extract local areas with irregular shapes and internal structural features from the 3D digital model as prototypes of irregular particles, separate them into independent particle models, and classify and store them in the ore digital twin library; Based on the independent particle models in the ore digital twin library, parameters are assigned to different mineral regions, the contact relationship of normal stiffness, tangential stiffness and friction coefficient parameters is defined, and a multiphase coupled particle model is generated.

[0008] As a preferred solution of the discrete element-based method for simulating the semi-autogenous grinding process in a mine, the method is as follows: based on the multiphase coupled particle model, the dynamic variable boundary surface is updated in the discrete element simulation framework, the particle motion trajectory and collision mode in the semi-autogenous grinding mill are tracked, and a simulation data set is generated, specifically: Import the multiphase coupled particle model into the discrete element simulation software, load the 3D geometric model of the semi-autogenous mill barrel, liner, inlet and outlet, and set the operating parameters; Enable the Bonded-Particle-Model contact model, set the weak bonding surface, and update the contact state to generate a particle motion trajectory data sequence; Based on the particle motion trajectory data sequence, the Hertz-Mindlin contact model is used to calculate the normal and tangential components of the contact force between particles. The collision frequency, collision energy peak and total energy dissipation are counted to generate a simulation data set.

[0009] As a preferred solution of the discrete element-based mining semi-autogenous grinding process simulation method of the present invention, the crushing characteristic label is obtained as follows: Based on the simulation data set, dynamic graph structure data is constructed with particles as nodes and contact forces as edges; A graph neural network model is used to process dynamic graph structure data, extract the contact force propagation path, and combine cluster analysis to generate a set of crushing characteristic labels related to different mineral areas.

[0010] As a preferred solution of the discrete element-based mining semi-autogenous grinding process simulation method of the present invention, the parameter mapping matrix is ​​generated as follows: Extract operating parameters and performance indicators based on simulation data sets and crushing characteristic label sets; The random forest regression algorithm is used to fit the mapping relationship between operating parameters and performance indicators to generate a parameter mapping matrix.

[0011] As a preferred solution of the discrete element-based mining semi-autogenous grinding process simulation method of the present invention, the particle size distribution prediction model is constructed as follows: Deploy IoT sensors to collect operational data, match it with the fragmentation feature label set, and generate a joint dataset; Based on the joint dataset, a long short-term memory network model was trained to generate the product particle size distribution curve and form a particle size distribution prediction model to predict the fine particle yield under different operating parameter combinations.

[0012] As a preferred solution of the discrete element-based mining semi-autogenous grinding process simulation method of the present invention, the energy-saving operation plan is obtained as follows: Obtain the time-of-use electricity price data of the power grid, establish a unit energy consumption cost series, and calculate the unit processing cost by combining the operating parameter combinations of the parameter mapping matrix and the fine particle yield under different operating parameter combinations; A genetic algorithm is used to search for the optimal operating parameter combination to maximize the fine particle yield and minimize the unit processing cost, generate a time-divided sequence of mill speed, filling rate and feed rate setting values, form an energy consumption optimization scheduling strategy, convert it into a control instruction sequence, and generate an energy-saving operation plan.

[0013] As a preferred solution of the discrete element-based mining semi-autogenous grinding process simulation method of the present invention, the generation of the calibration parameter set is specifically as follows: Execute energy-saving operation plans and collect actual operation data; The actual operation data are compared with the simulation data set, the deviation value sequence is calculated, and the normal stiffness, tangential stiffness and bond fracture energy parameters of the multiphase coupled particle model are adjusted until the deviation meets the process performance requirements, and a calibration parameter set is generated.

[0014] As a preferred solution of the discrete element-based mining semi-autogenous grinding process simulation method of the present invention, the construction of the grinding process perception mechanism and optimization control logic and the generation of a visual interface are specifically as follows: Based on the calibration parameter set, the control variable method is used to generate a performance indicator database, and an analysis report is generated through regression analysis; By integrating vibration sensor, acoustic emission sensor and lidar point cloud data, a grinding process perception mechanism is constructed to generate filling rate estimation and stubborn rock amount warning status, match the performance indicator database to form optimized control logic, and generate a visual interface.

[0015] In a second aspect, the present invention provides a discrete element-based mining semi-autogenous grinding process simulation system, comprising: The construction module collects the microscopic fracture network, bedding distribution, and mineral embedding characteristics of the ore sample. After generating the ore digital twin library through a 3D reconstruction algorithm, the prototype of the irregularly shaped particles is segmented and assigned with non-uniform physical property parameters to construct a multiphase coupled particle model. The mapping module, based on a multiphase coupled particle model, updates the dynamically variable boundary surface within a discrete element simulation framework to track particle motion trajectories and collision patterns within the SAG mill. After generating a simulation dataset, it uses a deep learning algorithm to mine the inter-particle contact force network topology and energy transfer paths, obtain crushing characteristic labels, and generate a parameter mapping matrix. The calibration module integrates IoT sensor data based on crushing characteristic labels and parameter mapping matrices to build a particle size distribution prediction model and energy consumption optimization scheduling strategy, obtain an energy-saving operation plan, compare simulation data with field production data, adjust the parameters of the multiphase coupled particle model, and generate a calibration parameter set. The analysis module uses the control variable method to study the influence of SAG mill operating parameters based on the calibration parameter set, obtains a performance indicator database and analysis report, and integrates multi-sensor data and lidar data to build a grinding process perception mechanism and optimize control logic, generating a visual interface. The summary module verifies the optimized control logic on site based on a visual interface, iteratively updates the multiphase coupled particle model and optimized control logic, and generates an implementation report.

[0016] The beneficial effects of the present invention are as follows: by fusing IoT sensor data with a set of crushing characteristic labels, a joint data set is generated; a long-short-term memory network model is used to train a particle size distribution prediction model; and by combining the grid's time-of-use electricity price and parameter mapping matrix, the operating parameters are optimized through a genetic algorithm to generate an energy-saving operation plan. This overcomes the limitations of static operating parameter settings, achieves parameter optimization under dynamic working conditions, significantly improves fine particle yield and energy efficiency, reduces equipment operating costs, and comprehensively improves simulation accuracy and process adaptability through multiphase modeling and data-driven optimization, providing an efficient and economical solution for the semi-autogenous grinding process in mines. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0018] Figure 1 This is a flow chart of the simulation method of the mining semi-autogenous grinding process based on discrete elements.

[0019] Figure 2 Build a flow chart for a multiphase coupled particle model.

[0020] Figure 3 Flowchart of simulation and optimization.

[0021] Figure 4 Flowchart of the sensing mechanism for the grinding process. DETAILED DESCRIPTION

[0022] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0023] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0024] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0025] Reference Figure 1, is an embodiment of the present invention, which provides a method for simulating a mining semi-autogenous grinding process based on discrete elements, comprising the following steps: S1: Collect the microscopic fracture network, bedding distribution and mineral embedding characteristics of the ore samples, generate the ore digital twin library through the 3D reconstruction algorithm, segment the irregular particle prototypes and assign non-uniform physical property parameters to construct a multiphase coupled particle model.

[0026] See also Figure 2 , specifically including the following steps: S1.1: Obtain representative ore samples from the mine site, cut and clean the ore samples, and prepare them into ore specimens suitable for industrial CT scanning; use an industrial CT scanner to perform a three-dimensional scan of the ore specimens and collect projection image data containing microscopic cracks, bedding structures, and mineral phase boundaries.

[0027] S1.2: Import the projection image data into the 3D reconstruction software to generate 3D volume data of the ore sample.

[0028] Median filtering is applied to the 3D volume data to reduce imaging noise and preserve the grayscale data of the internal structure. Based on the gradient differences of the grayscale data, a region growing segmentation algorithm is used to segment the 3D volume data to identify fracture areas, bedding planes, and areas of different mineral compositions. Apply the Marching-Cubes algorithm to mesh the surface of the segmented 3D volume data and generate a triangular facet model containing internal structure information; Smoothing and simplifying triangular patch models removes isolated noise points while preserving the geometric features of fracture networks, bedding planes, and mineral boundaries.

[0029] S1.3: Save the processed triangular face model in STL format to form a three-dimensional digital model of the ore.

[0030] Extracting a local area containing irregular shapes and internal structural features from the three-dimensional digital model as a prototype of the irregular particle; The subset extraction function of the 3D reconstruction software is used to separate the irregular particle prototypes into independent particle models. The surface details of each independent particle model are preserved to maintain the original surface roughness. All independent particle models are classified and stored to establish a digital twin library of ore.

[0031] S1.4: Based on the independent particle models in the ore digital twin library, the hardness parameters are assigned to the quartz-enriched area, sulfide area, and clay area according to the regional distribution of mineral composition within each independent particle model; based on the grayscale data in the three-dimensional volume data, a mapping relationship between grayscale data and mineral density is established, and the mineral density is assigned to the corresponding mineral area of ​​the independent particle model; based on the mineral type and fracture area characteristics, the fracture energy parameters are assigned to the quartz area, sulfide area, and fracture surface area of ​​the independent particle model; the independent particle model with assigned hardness parameters, density parameters, and fracture energy parameters is imported into the discrete element simulation software.

[0032] Through the material property configuration function of the discrete element simulation software, the hardness parameters, density parameters, and fracture energy parameters are bound to the corresponding geometric regions of the independent particle model; the normal stiffness parameters, tangential stiffness parameters, and friction coefficient parameters based on the interpolation of local physical properties define the contact relationship between different mineral regions within the independent particle model; the configured independent particle model and property parameters are saved to form a multiphase coupled particle model that can be used for simulation.

[0033] It is further explained that the gradual construction from projection image data to three-dimensional volume data, ore digital twin library and then to multiphase coupled particle model ensures smooth data transmission, lays a solid foundation for subsequent discrete element simulation, contact force analysis and process optimization, and helps to improve the efficiency of the semi-autogenous grinding process, reduce energy consumption and extend equipment life.

[0034] S2: Based on the multiphase coupled particle model, the dynamic variable boundary surface is updated in the discrete element simulation framework to track the particle motion trajectories and collision patterns in the semi-autogenous mill. After generating a simulation dataset, a deep learning algorithm is used to mine the contact force network topology and energy transfer paths between particles to obtain the crushing characteristic labels and parameter mapping matrix.

[0035] See also Figure 3 , specifically including the following steps: S2.1: Import the multiphase coupled particle model into the discrete element simulation software, load the three-dimensional geometric models of the SAG mill barrel, liner, and inlet and outlet, and form the simulation calculation domain; set the SAG mill operating parameters in the discrete element simulation software, including the preset speed, preset filling rate, and steel ball size according to the actual process ratio, for example, the speed is 75% of the critical speed, the filling rate is 30%, and the steel ball size is imported according to the actual process ratio; randomly place the multiphase coupled particle model into the SAG mill geometric model according to the actual ore-to-steel ball ratio to form the initial particle filling state.

[0036] Based on the initial particle filling state, the Bonded-Particle-Model contact model is enabled in the simulation domain, the crack area and bedding area inside the multiphase coupled particle model are set as weak bonding surfaces, and the bonding fracture energy threshold is defined; the simulation calculation is started, and the explicit time integration method is used to advance the simulation process with a preset time step; the contact state between the multiphase coupled particle model and the cylinder, liner and particles is updated in the simulation domain to determine whether a collision occurs and calculate the contact force; the position, velocity, acceleration and force vector of each particle in the simulation domain are recorded to generate a particle motion trajectory data sequence. Among them, the Bonded-Particle-Model contact model is a physical mechanics model based on Newtonian mechanics and contact mechanics principles, and does not require training.

[0037] Based on the particle trajectory data sequence, the normal and tangential components of the interparticle contact force are extracted, and the collision frequency, peak collision energy, and total energy dissipation are calculated. The process is as follows: Based on the particle trajectory data sequence, the contact point position and relative velocity of each pair of particles in the simulation domain are extracted. The Hertz-Mindlin contact model is used to calculate the normal contact force based on the position difference of the contact points. The tangential component is extracted based on the relative velocity of the particles at the contact points. The Mindlin-Deresiewicz theory is used to calculate the tangential contact force using the tangential stiffness and friction coefficient, combined with the tangential relative velocity, to generate the normal and tangential components of the interparticle contact force. Based on the normal and tangential components of the interparticle contact force, the number of collisions within each time step is counted, and the collision frequency is calculated. Based on the normal and tangential components of the interparticle contact force, the particle velocities before and after the collision are updated using Newton's second law and explicit time integration. The kinetic energy change is obtained by calculating the square difference between the particle mass and velocity, where the particle mass is derived from the density parameter of the multiphase coupled particle model; the maximum kinetic energy change is extracted as the collision energy peak, and the kinetic energy loss of all collisions is accumulated to generate the total energy dissipation; the broken particles are processed for sub-particle generation, and the geometric shape and physical property binding relationship of the broken particles are updated; the simulation is continuously run until the simulation calculation domain reaches a stable working condition, and the particle motion trajectory and collision mode data within the complete cycle are collected; the particle position, velocity, contact force and crushing event data are exported as a structured time series data set to form a simulation data set, where the Hertz-Mindlin contact model is a deterministic model based on mechanical formulas and does not require training.

[0038] S2.2: Based on the simulation dataset, a Python data analysis environment was imported to construct a dynamic graph structure data with particles as nodes and contact forces as edges. The process was as follows: the particle position, velocity, and inter-particle contact force within each time step were extracted to generate a particle attribute dataset. Each particle was defined as a node in the dynamic graph structure data, and the contact force relationships between particles were defined as edges. The initial graph structure was generated, and the vector sum of the normal and tangential components of the inter-particle contact force was calculated as the edge weight. The initial graph structure was updated to generate the dynamic graph structure data. The particle position and velocity of the dynamic graph structure data were used as node features, and the contact force magnitude was used as the edge feature. A graph neural network model was used for training. The node features were updated through multi-layer graph convolution operations. The contact force information of adjacent particles was aggregated to extract the propagation path and aggregation pattern of the contact force in the dynamic graph structure data.

[0039] Cluster analysis is performed on the output features of the graph neural network to classify particles with similar contact force propagation behaviors into the same category. Based on the clustering results, crushing behavior category labels for different ore types are generated to form a crushing characteristic label set, specifically: Based on the output features of the graph neural network model, the contact force propagation path and clustering pattern characteristics of each particle are extracted, generating a set of particle feature vectors. Cluster analysis is then performed using the K-Means clustering algorithm: Based on the particle feature vectors output by the graph neural network model, the contact force propagation path and clustering pattern characteristics of each particle are used as input data. Multiple cluster centers are initialized. The Euclidean distance between each particle feature vector and the cluster center is iteratively calculated, and the particle is assigned to the cluster center with the closest distance. The cluster center is then updated to the mean of the assigned particle's feature vector. This iteration is repeated until the cluster center stabilizes or a preset number of iterations is reached. Particles are then divided into multiple categories based on the similarity of their contact force propagation behavior, generating particle cluster categories. The ore type distribution of particles within each cluster category is analyzed, and crushing behavior characteristics associated with quartz-rich regions, sulfide regions, and clay regions are extracted to generate crushing behavior category labels for different ore types. All crushing behavior category labels are categorized and stored as a structured label dataset, forming a crushing characteristic label set.

[0040] S2.3: Classify the simulation operating condition data under different operating parameter combinations, establish the corresponding relationship between the SAG mill operating parameters and the crushing characteristic label set, and calculate the fine particle generation rate, crushing energy consumption and stubborn stone ratio under each set of SAG mill operating parameters to form a multidimensional response index set; use the SAG mill operating parameters as input variables and the multidimensional response index set as output variables to fit the nonlinear mapping relationship between parameters and performance; save the mapping relationship model and coefficient configuration to form a parameter mapping matrix, specifically: (1) Based on the operating parameters of the semi-autogenous grinding mill, the rotational speed, filling rate and steel ball size were extracted as input variables to generate an operating parameter data set. Based on the multidimensional response index set, the fine particle generation rate, crushing energy consumption and stubborn stone ratio were extracted as output variables to generate a response index data set.

[0041] (2) Based on the operating parameter data set and the response index data set, the speed, filling rate and steel ball size of the semi-autogenous mill operating parameters and the fine particle generation rate, crushing energy consumption and stubborn stone ratio of the multidimensional response index set are normalized. After generating the standardized operating parameter data set and the standardized response index data set, the standardized operating parameter data set and the standardized response index data set are divided into a training data set and a validation data set. The training data set is used to construct the decision tree of the random forest regression algorithm, and the validation data set is used to evaluate the performance of the mapping relationship model. Based on the training data set, the random forest regression algorithm is used to construct multiple decision trees by randomly sampling a subset of the semi-autogenous mill operating parameters. Each decision tree is based on the speed, The filling rate and steel ball size are used to predict the fine particle generation rate, crushing energy consumption and stubborn stone ratio. After generating a decision tree set, the prediction output of the random forest regression algorithm is generated by averaging the prediction results of each decision tree, fitting the nonlinear mapping relationship between the semi-autogenous mill operating parameters and the multidimensional response index set, and generating a mapping relationship model. Based on the validation data set, the error between the prediction output of the mapping relationship model and the actual multidimensional response index set is calculated, the number of trees and the feature selection ratio of the decision tree set are adjusted, and the prediction accuracy of the mapping relationship model is optimized. Based on the optimized mapping relationship model, it is verified that the prediction error on the validation data set meets the preset accuracy requirements. The final decision tree set and weight configuration are saved to generate a mapping relationship model.

[0042] (3) Based on the mapping relationship model, the tree structure weights and feature importance coefficients of the random forest regression algorithm are extracted to generate a coefficient configuration data set. The weights and coefficients of the mapping relationship model are organized into structured data and saved as a parameter mapping matrix.

[0043] It is further explained that by simulating the crushing behavior of different ore types and optimizing operating parameters, a crushing characteristic label set and parameter mapping matrix for mineral areas such as quartz, sulfide and clay are generated. This can adapt to a variety of ore types and process conditions, optimize the fine particle generation rate, reduce crushing energy consumption, and reduce the proportion of stubborn stone, thereby improving the efficiency of the semi-autogenous grinding process, reducing energy consumption and extending the service life of the equipment.

[0044] S3: Based on the crushing characteristic labels and parameter mapping matrix, IoT sensor data is integrated to build a particle size distribution prediction model and energy consumption optimization scheduling strategy to obtain an energy-saving operation plan. The simulation dataset is compared with the on-site production data, the multiphase coupled particle model parameters are adjusted, and a calibration parameter set is generated.

[0045] The specific steps include: S3.1: Deploy IoT sensors, including power sensors, vibration sensors, slurry concentration meters, and feed weighing devices, at the SAG mill site to collect operating data in real time. Synchronize the operating data collected by the IoT sensors to the industrial data platform in a timestamp-aligned manner to form a continuous operating parameter time series stream. Perform spatiotemporal matching between the operating parameter time series stream and the crushing characteristic label set to construct a joint dataset containing operating condition characteristics and crushing behavior. The process is as follows: Based on the operating parameter time series stream, extract the SAG mill operating parameters within each time step, including speed, fill rate, feed rate, and slurry concentration, to form an operating parameter feature sequence. Based on the crushing characteristic label set, extract the crushing behavior category labels for different ore types corresponding to the timestamps of the operating parameter time series stream to form a crushing behavior label sequence. Based on the timestamps, align and match the operating parameter feature sequence with the crushing behavior label sequence to generate time-synchronized operating condition-behavior data pairs. Through the time-synchronized operating condition-behavior data pairs, integrate the SAG mill operating parameters and crushing behavior category labels to generate a joint dataset containing operating condition characteristics and crushing behavior.

[0046] S3.1.1: Train a long-short-term memory network model based on the combined dataset, process the time series data through a multi-layer network structure, predict the product particle size distribution curve, optimize the model weights to form a particle size distribution prediction model, and predict the fine particle yield under different operating parameter combinations. Specifically: (1) Based on the joint data set, the time-synchronized SAG mill operating parameters and crushing behavior category labels are extracted to construct serialized input data containing time steps; based on the serialized input data, a long short-term memory network model is used for training, and the input layer is set to receive the SAG mill operating parameters and crushing behavior category labels, the hidden layer processes the time series features, and the output layer generates the product particle size distribution curve; based on the predicted particle size distribution curve of the long short-term memory network model and the laboratory screening results, the mean square error is calculated as the prediction error to measure the deviation of the particle size distribution curve, and the gradient descent method is used to adjust the weight of the long short-term memory network model. When the mean square error is lower than the preset error threshold, it indicates convergence and a particle size distribution prediction model is generated. The error threshold is customized based on the required prediction accuracy.

[0047] (2) Based on the particle size distribution prediction model, different combinations of SAG mill operating parameters are input to predict the corresponding product particle size distribution curve, from which the proportion of particles in the target particle size range is extracted to generate the fine particle yield.

[0048] S3.1.2: Obtain time-of-use electricity price data from the power grid, divide the time periods into peak, flat, and valley periods, establish unit energy consumption cost series for different time periods, and combine the unit energy consumption cost series with the fine particle yield of the particle size distribution prediction model to form an energy-saving operation plan.

[0049] The operating parameter combinations in the parameter mapping matrix are associated with the corresponding energy consumption indicators, combined with the unit energy consumption cost sequence and the predicted fine particle yield. A genetic algorithm is used to search for the optimal operating parameter combination within the parameter mapping matrix to maximize the fine particle yield and minimize the unit processing cost. A time-segmented sequence of mill speed, fill rate, and feed rate setpoints is generated to form an energy-optimized scheduling strategy. The energy-optimized scheduling strategy is converted into an executable control instruction sequence to form an energy-saving operation plan, specifically: (1) Based on the parameter mapping matrix, the semi-autogenous mill operating parameter combinations, including speed, filling rate and feed rate, and the corresponding energy consumption indicators, are extracted to generate an operating parameter-energy consumption data set. Based on the operating parameter-energy consumption data set and combined with the unit energy consumption cost sequence, the unit processing cost of each operating parameter combination is calculated; based on the particle size distribution prediction model, the predicted fine particle yield is extracted, and combined with the unit processing cost, a weighted fitness function is constructed, which assigns a positive weight to the fine particle yield and a negative weight to the unit processing cost, and comprehensively evaluates the advantages and disadvantages of the operating parameter combination; based on the weighted fitness function, a genetic algorithm is used to initialize a population containing multiple operating parameter combinations, and the population is iteratively updated through selection, crossover and mutation operations. The fitness value of each operating parameter combination is calculated, and the operating parameter combination with the highest fine particle yield and the lowest unit processing cost is screened out. When the fitness value converges, the optimal operating parameter combination is output.

[0050] (2) Based on the optimal operating parameter combination obtained by genetic algorithm search, a time-divided sequence of mill speed, filling rate and feed rate setting values ​​is generated according to the peak, flat and valley periods of the grid time-of-use electricity price, forming an energy consumption optimization scheduling strategy. According to the energy consumption optimization scheduling strategy, the time-divided sequence of setting values ​​is converted into a control instruction sequence that can be recognized by the semi-autogenous mill control equipment, forming an energy-saving operation plan.

[0051] S3.2: Execute the energy-saving operation plan in the mine production center, and simultaneously collect the particle size distribution, power consumption and liner wear data in actual operation to form on-site production data; compare the on-site production data with the simulation results of the corresponding working conditions in the simulation data set item by item, and calculate the deviation value sequence of fine particle yield, peak power and stone ratio. The process is as follows: based on the on-site production data, extract the fine particle yield, peak power and stone ratio under the actual operating conditions to generate an on-site performance index sequence; based on the simulation data set, extract the simulation conditions with the same operating parameter combination as the on-site production data, obtain the corresponding fine particle yield, peak power and stone ratio, and generate a simulation performance index sequence; through the on-site performance index sequence and the simulation performance index sequence, calculate the absolute difference of fine particle yield, peak power and stone ratio for each time step, generate a single deviation value, and sort it by time step, integrate the deviations of fine particle yield, peak power and stone ratio, and generate a time-series deviation value sequence.

[0052] Based on the statistical trend of the deviation value sequence, the regular errors of the normal stiffness, tangential stiffness and bond fracture energy parameters in the multiphase coupled particle model are identified; the normal stiffness, tangential stiffness and bond fracture energy parameters are adjusted in sequence in the discrete element simulation software, and the corresponding working condition simulation is re-run; the simulation results after adjusting the parameters are compared again with the on-site production data to verify the convergence of the deviation value sequence.

[0053] When the deviations of fine particle yield, peak power, and coarse stone ratio are all below the parameter threshold, the current parameter configuration is saved. The parameter threshold is set based on the performance requirements of the semi-autogenous mill process; the saved normal stiffness, tangential stiffness, and bond fracture energy parameter combination is defined as the calibration parameter set.

[0054] Preferably, by combining data sets, particle size distribution prediction models and energy consumption optimization scheduling strategies, different ore types (such as quartz, sulfide) and operating conditions (such as peak and valley electricity prices) are adapted to optimize fine particle yield, reduce energy consumption and liner wear; by calibrating parameter sets to ensure that simulation is consistent with actual production, equipment overload and wear are reduced; compared with existing optimization methods for a single operating condition, the multi-condition adaptability of step S3 significantly improves the operating efficiency and equipment service life of the semi-autogenous grinding mill.

[0055] S4: Based on the calibration parameter set, the control variable method is used to study the influence of the semi-autogenous grinding mill operating parameters, obtain a performance indicator database and analysis report, and integrate multi-sensor data and lidar data to build a grinding process perception mechanism and optimize the control logic to generate a visual interface.

[0056] The specific steps include: S4.1: Based on the calibration parameter set, the control variable method is used to study the influence of the SAG mill operating parameters and generate a performance index database and analysis report. Specifically: The calibration parameter set was loaded into the discrete element simulation software as the baseline model parameters for the simulation experiment. The SAG mill speed, filling rate, steel ball ratio and liner lift angle were set as variables in sequence. For each variable, multiple different values ​​were set, while other parameters remained unchanged. The simulation was run, and the fine particle output, unit energy consumption, liner wear rate and stone accumulation under each operating condition were recorded. The recorded data under all operating conditions were summarized to establish a performance indicator database containing the correspondence between operating parameters and performance indicators. Trend analysis was performed on the data in the performance indicator database to generate an analysis report on the influence of SAG mill operating parameters. The trend analysis process is as follows: based on the performance index database, data pairs of SAG mill operating parameters (including speed, filling rate, steel ball ratio and liner lift angle) and corresponding performance indicators (including fine particle output, specific energy consumption, liner wear rate and stubborn stone accumulation) are extracted; for each operating parameter, regression analysis is performed on fine particle output, specific energy consumption, liner wear rate and stubborn stone accumulation, and the functional relationship between each operating parameter and performance indicator is fitted; based on the fitted functional relationship, the interpolation method is used to generate continuous data points, and trend curves of speed, filling rate, steel ball ratio and liner lift angle with each performance indicator are plotted.

[0057] S4.2: Based on the performance indicator database and multi-sensor data, integrate the lidar data, build the grinding process perception mechanism and optimize the control logic, and generate a visual interface. Figure 4 , specifically: (1) Vibration sensors and acoustic emission sensors are installed on the outside of the semi-autogenous grinding mill barrel, and a lidar scanner is arranged at the inlet and outlet ends to synchronously collect vibration signals, acoustic wave signals and lidar point cloud data. The lidar point cloud data reflects the spatial position and trajectory of the material.

[0058] The vibration signal and acoustic wave signal are transformed into time-frequency to extract the energy amplitude of the characteristic frequency band; the lidar point cloud data is reconstructed in three dimensions to generate a dynamic image of the material distribution and movement state inside the mill; the energy amplitude of the characteristic frequency band is correlated with the material filling height and falling trajectory in the dynamic image to establish an empirical correspondence between the vibration energy amplitude and material filling rate, and between the acoustic wave signal characteristics and the amount of stubborn stone accumulation. These relationships are encoded into state recognition rules to form a perception mechanism for the grinding process.

[0059] The specific process for conducting correlation analysis and forming a perception mechanism for the grinding process involves using correlation analysis to calculate the correlation coefficient between the energy amplitude of the characteristic frequency band and the material filling height and drop trajectory. The process involves converting the time-domain signal data collected by the vibration sensor and acoustic emission sensor into the frequency domain through a fast Fourier transform, extracting the energy amplitude of the characteristic frequency band, and generating a characteristic frequency band energy amplitude sequence. The geometric distribution of the material within the semi-autogenous grinding mill is reconstructed based on LiDAR point cloud data, and the time series data of the material filling height and drop trajectory are calculated to generate a material filling height sequence and a drop trajectory sequence. The Pearson correlation coefficient method was used to calculate the correlation coefficients between the energy amplitude sequence of the characteristic frequency band and the material filling height sequence and the dropping trajectory sequence, respectively, to quantify the linear correlation strength between the vibration energy amplitude and the material filling height and the dropping trajectory. Based on the correlation analysis results, regression analysis was used to fit the functional relationship between the vibration energy amplitude and the material filling rate, and between the acoustic signal characteristics and the amount of rock accumulation, and to establish an empirical correspondence. The empirical correspondence was converted into a conditional judgment rule, and the threshold range of the vibration energy amplitude and the acoustic signal characteristics was set, corresponding to the state interval of the material filling rate and the amount of rock accumulation, and encoded as a state recognition rule. Based on the state recognition rule, a grinding process perception mechanism was constructed to identify the material filling rate and rock accumulation status in real time during the operation of the semi-autogenous grinding mill.

[0060] To convert the empirical correspondence into a conditional judgment rule, set the threshold range of the vibration energy amplitude and the acoustic signal characteristics, and correspond to the state interval of the material filling rate and the amount of stubborn stone accumulation. The following is an example of encoding the state recognition rule: Based on the empirical correspondence, it is shown that when the vibration energy amplitude is higher, the material filling rate is higher, and when the intensity of the specific frequency band of the acoustic signal feature is higher, the amount of stubborn rock accumulation is larger. Based on this, the conditional judgment rule is set as follows: if the vibration energy amplitude is greater than the vibration feature threshold, the material filling rate is judged to be in a high state, otherwise, the material filling rate is judged to be in a low state. If the intensity of the specific frequency band of the acoustic signal feature is greater than the acoustic feature threshold, the amount of stubborn rock accumulation is judged to be in a high state, otherwise, the amount of stubborn rock accumulation is judged to be in a low state. These conditional judgment rules are encoded as state recognition rules. For example, when the vibration energy amplitude exceeds the vibration feature threshold range and the intensity of the specific frequency band of the acoustic signal feature is lower than the acoustic feature threshold range, it is identified as an operating state of "high material filling rate and low stubborn rock accumulation"; when the vibration energy amplitude is lower than the vibration feature threshold range and the intensity of the specific frequency band of the acoustic signal feature exceeds the acoustic feature threshold range, it is identified as an operating state of "low material filling rate and high stubborn rock accumulation". The vibration characteristic threshold range and the acoustic wave characteristic threshold range are set based on statistical analysis of empirical correspondences. Based on the vibration characteristic threshold range and the acoustic wave characteristic threshold range, a state recognition rule is generated to identify the material filling rate and the stubborn stone accumulation state of the semi-autogenous grinding mill in real time.

[0061] (2) Match the optimal operating condition combination in the performance index database with the state recognition rules of the grinding process perception mechanism. When the grinding process perception mechanism identifies that the filling rate is low and the amount of stubborn stones increases, it is set to trigger the response action of increasing the feed rate and adjusting the rotation speed. When the grinding process perception mechanism identifies that the energy consumption increases abnormally, it is set to trigger the response action of reducing the filling rate or adjusting the steel ball ratio. The response actions are combined into a control instruction sequence to form an optimized control logic.

[0062] The human-computer interaction software is imported into the mill's three-dimensional model and real-time sensor data. The software interface displays the material movement status image, fill rate estimation, stubborn stone quantity warning status and recommended control instructions in real time. The control instructions are connected to the on-site programmable logic controller to realize operation suggestion push and remote parameter setting, generating a visual interface.

[0063] It is further explained that the empirical correspondence between vibration energy amplitude and material filling rate, acoustic signal characteristics and stubborn stone accumulation amount is established, encoded into state recognition rules, and the process of constructing the grinding process perception mechanism is compared with the process perception method in the prior art that often relies on a single sensor or static monitoring. The present invention comprehensively characterizes the material movement state inside the mill through multi-source data fusion (vibration, acoustic wave and lidar point cloud) and dynamic three-dimensional reconstruction, significantly improving the accuracy and real-time performance of operating status perception.

[0064] S5: Based on the visual interface, verify the optimized control logic on site, iteratively update the multiphase coupled particle model and optimized control logic, and generate an implementation report.

[0065] The specific steps include: A visualization interface generated by S4.2 displays the SAG mill's material movement state image, fill rate estimate, rock content warning status, and recommended control instructions in real time. The fill rate estimate and rock content warning status are generated using the state recognition rules of the grinding process perception mechanism and inferred based on actual operating data collected by vibration sensors, acoustic emission sensors, and lidar scanners. A control instruction sequence for the optimized control logic is executed at the mine site to collect actual operating data, including fine particle production, specific energy consumption, and liner wear rate. The fine particle production, specific energy consumption, and liner wear rate in the actual operating data are compared with the fill rate estimate and rock content warning status displayed on the visualization interface, and the deviation is calculated. Based on the deviation, the physical property parameters of the multiphase coupled particle model (such as normal stiffness, tangential stiffness, and bond fracture energy) and the response actions of the optimized control logic (such as feed rate or speed adjustment) are adjusted. Field verification and parameter adjustments are repeated until the deviation meets the SAG mill's process performance requirements.

[0066] Summarize the verification results, parameter adjustment records and update plans for optimized control logic, generate implementation reports, and record process optimization effects and operating parameter configurations.

[0067] This embodiment also provides a mine semi-autogenous grinding process simulation system based on discrete elements, including: The construction module collects the microscopic fracture network, bedding distribution and mineral embedding characteristics of the ore samples, generates the ore digital twin library through the three-dimensional reconstruction algorithm, divides the irregular particle prototypes and assigns non-uniform physical property parameters to construct a multiphase coupled particle model.

[0068] The mapping module, based on a multiphase coupled particle model, updates the dynamically variable boundary surface in a discrete element simulation framework, tracks the motion trajectories and collision patterns of particles in the semi-autogenous mill, generates a simulation dataset, and uses a deep learning algorithm to mine the contact force network topology and energy transfer paths between particles, obtains the crushing characteristic labels, and generates a parameter mapping matrix.

[0069] The calibration module integrates IoT sensor data based on the crushing characteristic labels and parameter mapping matrix to build a particle size distribution prediction model and energy consumption optimization scheduling strategy, obtain an energy-saving operation plan, compare the simulation data with the on-site production data, adjust the parameters of the multiphase coupled particle model, and generate a calibration parameter set.

[0070] The analysis module uses the control variable method to study the influencing patterns of the semi-autogenous grinding mill operating parameters based on the calibration parameter set, obtains a performance indicator database and analysis report, and integrates multi-sensor data and lidar data to build a grinding process perception mechanism and optimize control logic, generating a visual interface.

[0071] The summary module verifies the optimized control logic on site based on a visual interface, iteratively updates the multiphase coupled particle model and optimized control logic, and generates an implementation report.

[0072] This embodiment also provides a computer device, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute computer-executable instructions to implement the discrete element-based mining semi-autogenous grinding process simulation method proposed in the above embodiment.

[0073] The computer device may be a terminal, comprising a processor, memory, a communication interface, a display, and an input device connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores an operating system and computer programs. The internal memory provides an environment for the operating system and computer programs stored in the non-volatile storage media. The communication interface of the computer device is used to communicate with external terminals via wired or wireless communication. Wireless communication may be achieved via Wi-Fi, a carrier network, NFC (near-field communication), or other technologies. The display of the computer device may be a liquid crystal display or an electronic ink display. The input device may be a touchscreen overlay on the display, buttons, a trackball, or a touchpad on the computer device housing, or an external keyboard, touchpad, or mouse.

[0074] This embodiment also provides a storage medium having a computer program stored thereon, which, when executed by a processor, implements the discrete element-based mining semi-autogenous grinding process simulation method proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0075] In summary, the present invention generates a joint data set by integrating IoT sensor data with a set of crushing characteristic labels, adopts a long short-term memory network model to train a particle size distribution prediction model, combines the grid time-of-use electricity price and parameter mapping matrix, optimizes operating parameters through a genetic algorithm, and generates an energy-saving operation plan. It overcomes the limitations of static operating parameter settings, realizes parameter optimization under dynamic working conditions, significantly improves fine particle yield and energy efficiency, reduces equipment operating costs, and comprehensively improves simulation accuracy and process adaptability through multiphase modeling and data-driven optimization, providing an efficient and economical solution for the semi-autogenous grinding process in mines.

[0076] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for simulating a mining semi-autogenous grinding process based on discrete elements, characterized by: include: The microscopic fracture network, bedding distribution, and mineral embedding characteristics of the ore samples are collected. A digital twin library of the ore is generated through a 3D reconstruction algorithm. The prototypes of the irregular particles are then segmented and assigned non-uniform physical property parameters to construct a multi-phase coupled particle model. Based on a multiphase coupled particle model, a dynamic variable boundary surface is updated in a discrete element simulation framework to track the particle motion trajectories and collision patterns within the semi-autogenous mill. After generating a simulation dataset, a deep learning algorithm is used to mine the inter-particle contact force network topology and energy transfer paths, obtain the crushing characteristic labels, and generate a parameter mapping matrix. Based on the crushing characteristic labels and parameter mapping matrix, IoT sensor data is integrated to build a particle size distribution prediction model and energy consumption optimization scheduling strategy to obtain an energy-saving operation plan. The simulation data set is compared with the field production data, and the parameters of the multiphase coupled particle model are adjusted to generate a calibration parameter set. Based on the calibration parameter set, the control variable method is used to study the influence of the semi-autogenous grinding mill operating parameters, obtain a performance indicator database and analysis report, and integrate multi-sensor data and lidar data to build a grinding process perception mechanism and optimize control logic, generating a visual interface. Based on the visual interface, the optimized control logic is verified on site, the multiphase coupled particle model and optimized control logic are iteratively updated, and an implementation report is generated.

2. The method for simulating a mining semi-autogenous grinding process based on discrete element method according to claim 1, characterized in that: The microscopic fracture network, bedding distribution and mineral embedding characteristics of the ore sample are collected, and the ore digital twin library is generated through a three-dimensional reconstruction algorithm. After that, the prototype of the irregular particles is segmented and assigned with non-uniform physical property parameters to construct a multi-phase coupled particle model. Specifically, Obtain representative ore samples from the mine site, perform 3D scanning, and collect projection image data including microscopic cracks, bedding structures, and mineral phase boundaries; Import the projection image data into the 3D reconstruction software to generate 3D volume data, and then form a 3D digital model through preprocessing; Extract local areas with irregular shapes and internal structural features from the 3D digital model as prototypes of irregular particles, separate them into independent particle models, and classify and store them in the ore digital twin library; Based on the independent particle models in the ore digital twin library, parameters are assigned to different mineral regions, the contact relationship of normal stiffness, tangential stiffness and friction coefficient parameters is defined, and a multiphase coupled particle model is generated.

3. The method for simulating a mining semi-autogenous grinding process based on discrete element method according to claim 2, characterized in that: Based on the multiphase coupled particle model, the dynamic variable boundary surface is updated in the discrete element simulation framework to track the particle motion trajectory and collision mode in the semi-autogenous mill and generate a simulation data set, specifically: Import the multiphase coupled particle model into the discrete element simulation software, load the 3D geometric model of the semi-autogenous mill barrel, liner, inlet and outlet, and set the operating parameters; Enable the Bonded-Particle-Model contact model, set the weak bonding surface, and update the contact state to generate a particle motion trajectory data sequence; Based on the particle motion trajectory data sequence, the Hertz-Mindlin contact model is used to calculate the normal and tangential components of the contact force between particles. The collision frequency, collision energy peak and total energy dissipation are counted to generate a simulation data set.

4. The method for simulating a mining semi-autogenous grinding process based on discrete element method according to claim 3, characterized in that: The fragmentation characteristic label is obtained as follows: Based on the simulation data set, dynamic graph structure data is constructed with particles as nodes and contact forces as edges; A graph neural network model is used to process dynamic graph structure data, extract the contact force propagation path, and combine cluster analysis to generate a set of crushing characteristic labels related to different mineral areas.

5. The method for simulating a mining semi-autogenous grinding process based on discrete element method according to claim 4, characterized in that: The generation parameter mapping matrix is ​​specifically: Extract operating parameters and performance indicators based on simulation data sets and crushing characteristic label sets; The random forest regression algorithm is used to fit the mapping relationship between operating parameters and performance indicators to generate a parameter mapping matrix.

6. The method for simulating a mining semi-autogenous grinding process based on discrete element method according to claim 5, characterized in that: The particle size distribution prediction model is constructed as follows: Deploy IoT sensors to collect operational data, match it with the fragmentation feature label set, and generate a joint dataset; Based on the joint dataset, a long short-term memory network model was trained to generate the product particle size distribution curve, forming a particle size distribution prediction model to predict the fine particle yield under different operating parameter combinations.

7. The method for simulating a mining semi-autogenous grinding process based on discrete element method according to claim 6, characterized in that: The energy-saving operation plan is specifically as follows: Obtain the time-of-use electricity price data of the power grid, establish a unit energy consumption cost series, and calculate the unit processing cost by combining the operating parameter combinations of the parameter mapping matrix and the fine particle yield under different operating parameter combinations; A genetic algorithm is used to search for the optimal operating parameter combination to maximize the fine particle yield and minimize the unit processing cost, generate a time-divided sequence of mill speed, filling rate and feed rate setting values, form an energy consumption optimization scheduling strategy, convert it into a control instruction sequence, and generate an energy-saving operation plan.

8. The method for simulating a mining semi-autogenous grinding process based on discrete element method according to claim 7, characterized in that: The generating of the calibration parameter set is specifically as follows: Execute energy-saving operation plans and collect actual operation data; The actual operation data are compared with the simulation data set, the deviation value sequence is calculated, and the normal stiffness, tangential stiffness and bond fracture energy parameters of the multiphase coupled particle model are adjusted until the deviation meets the process performance requirements, and a calibration parameter set is generated.

9. The method for simulating a mining semi-autogenous grinding process based on discrete element method according to claim 8, characterized in that: The construction of the grinding process perception mechanism and optimization control logic generates a visual interface, specifically: Based on the calibration parameter set, the control variable method is used to generate a performance indicator database, and an analysis report is generated through regression analysis; By integrating vibration sensor, acoustic emission sensor and lidar point cloud data, a grinding process perception mechanism is constructed to generate filling rate estimation and stubborn rock amount warning status, match the performance indicator database to form optimized control logic, and generate a visual interface.

10. A mine semi-autogenous grinding process simulation system based on discrete element method, based on the mine semi-autogenous grinding process simulation method based on discrete element method according to any one of claims 1 to 9, characterized in that: include: The construction module collects the microscopic fracture network, bedding distribution, and mineral embedding characteristics of the ore sample. After generating the ore digital twin library through a 3D reconstruction algorithm, the prototype of the irregularly shaped particles is segmented and assigned with non-uniform physical property parameters to construct a multiphase coupled particle model. The mapping module, based on a multiphase coupled particle model, updates the dynamically variable boundary surface within a discrete element simulation framework to track particle motion trajectories and collision patterns within the SAG mill. After generating a simulation dataset, it uses a deep learning algorithm to mine the inter-particle contact force network topology and energy transfer paths, obtain crushing characteristic labels, and generate a parameter mapping matrix. The calibration module integrates IoT sensor data based on crushing characteristic labels and parameter mapping matrices to build a particle size distribution prediction model and energy consumption optimization scheduling strategy, obtain an energy-saving operation plan, compare simulation data with field production data, adjust the parameters of the multiphase coupled particle model, and generate a calibration parameter set. The analysis module uses the control variable method to study the influence of SAG mill operating parameters based on the calibration parameter set, obtains a performance indicator database and analysis report, and integrates multi-sensor data and lidar data to build a grinding process perception mechanism and optimize control logic, generating a visual interface. The summary module verifies the optimized control logic on site based on a visual interface, iteratively updates the multiphase coupled particle model and optimized control logic, and generates an implementation report.

Images