Intelligent analysis and optimization method and system for running state of coal mine conveyor

Abstract

The invention belongs to the technical field of coal mines, and discloses an intelligent analysis and optimization method and system for the running state of a coal mine conveyor. The method comprises the following steps: inputting the bandwidth width, the adhesive tape type, the carrier roller specification and the like of a transfer point of a hopper of a belt conveyor, physical attribute parameters of particle materials and environmental humidity parameters into Bentley software, and constructing a three-dimensional geometric model; the local adhesion characteristics of the discrete units are dynamically updated through moisture film migration; dynamically distributing the friction coefficient of the boundary area according to the contact frequency and the pressure; constructing a space-time intersection grid to capture movement track intersection characteristics; determining a stacking boundary curved surface through fracture and reconstruction of the stress transfer chain; the curvature change rate is analyzed to determine an actual discharging angle; and constructing an adjustment matrix iterative optimization model based on the deviation with the target unloading angle. According to the method, various permutations and combinations of assemblies such as different driving modes, head and tail frames and tensioning devices can be processed, the modeling precision and optimization efficiency are improved, and the adaptability of the system to complex working conditions is enhanced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of coal mines, more particularly, the present application relates to a coal mine conveyor operation state intelligent analysis and optimization method and system. BACKGROUND

[0002] Coal, as a non-renewable resource, has always been one of the important mining and burning resources in China. With the continuous development of coal mining technology and the improvement of automation level, the belt conveyor, as a kind of efficient and continuous coal transportation equipment, has become the core equipment of the modern underground and ground transportation system in coal mines, and is widely used in key links such as coal mining face, main transportation roadway, coal preparation plant and coal storage yard, for conveying different particle size and humidity of coal materials such as raw coal, clean coal, medium coal and coal gangue.

[0003] However, the existing technology has obvious deficiencies in simulating the flow behavior of coal particle materials at the transfer point of the hopper, especially in dealing with the change of flow characteristics of coal materials under complex environmental conditions in the mine. For example, in the actual coal mine production scene, the environmental humidity of the underground fully mechanized mining face can reach more than 85%, while the humidity of the ground coal storage yard can exceed 90% in the rainy season, and may be as low as 25% below in the dry season. The viscosity characteristics of coal particles (such as raw coal with particle size of 0.5-50mm) will change significantly due to such humidity fluctuations, especially high-moisture lignite and long-flame coal are more likely to stick together. The existing technology usually uses static coal material attribute parameters and simplified boundary conditions for modeling, ignoring the dynamic influence of mine humidity change on the surface water film of coal particles, failing to accurately capture the migration path and migration speed change of the water film on the surface of coal particles, resulting in serious inconsistency between the local adhesion characteristics of coal materials in the simulation and the actual underground working conditions. In addition, the existing technology often assumes that the friction coefficient is a fixed value in the boundary surface treatment of the hopper transfer point, and fails to adaptively adjust according to the dynamic changes of the contact frequency and contact pressure of coal materials and the boundary surface, especially under high humidity conditions in the mine, wet coal is easy to form an adhesion layer on the hopper wall, causing problems such as blockage and coal piling, which seriously affect the continuity of coal mine production.

[0004] The prior art lacks a refined analysis of the intersection characteristics of the coal particle motion trajectory, the dynamic fracture and reconstruction process of the stress transmission chain, and is difficult to accurately describe the local flow field behavior of the coal material inside the hopper transfer point and the formation mechanism of the accumulation boundary surface. This rough simulation method directly leads to a large deviation between the actual discharge angle and the model predicted value. In the actual application of the coal mine belt conveyor, the system needs to handle the complex combination of various component parameters, including adapting to different coal mining faces, belt width, wear-resistant flame-retardant belt types, underground explosion-proof roller types, variable frequency drive forms, head and tail frame types suitable for roadway conditions, and automatic tensioning devices. The existing modeling technology is difficult to effectively integrate the comprehensive influence of these parameter changes on the flow behavior of coal materials, resulting in a large difference between the optimization results and the actual working conditions of the coal mine, affecting the efficiency and safety of the coal mine transportation operation.

[0005] In view of this, the present application provides a coal mine conveyor operation state intelligent analysis and optimization method and system to solve the above problems. SUMMARY

[0006] In order to overcome the above-mentioned defects of the prior art, in order to achieve the above-mentioned purpose, on the one hand, the present application provides a coal mine conveyor operation state intelligent analysis and optimization method, comprising: Inputting the geometric structure parameters of the belt conveyor hopper transfer point, the physical property parameters of the particle material, and the environmental humidity parameters in the Bentley software, the physical property parameters including the particle size distribution, density and initial viscosity of the particle material, and constructing a three-dimensional geometric model of the hopper transfer point based on the geometric structure parameters; Discretize the physical property parameters of the particle material into a plurality of discrete units to form an initial particle material discrete unit set, and map the discrete unit set into the three-dimensional geometric model; Divide the surface of each discrete unit into a plurality of micro-grid regions, and dynamically update the local adhesion characteristics of each discrete unit by the change of the migration path and migration speed of the water film between the micro-grid regions under different environmental humidity parameters; Divide the boundary surface of the hopper transfer point into a plurality of local boundary regions, and dynamically assign the friction coefficient of each local boundary region according to the contact frequency and contact pressure of each local boundary region with the particle material discrete unit set; Based on the updated local adhesion characteristics and the dynamically assigned friction coefficient, simulate the motion process of the particle material discrete unit set in the three-dimensional geometric model, capture the intersection characteristics of the discrete unit motion trajectory inside the hopper transfer point by constructing a space-time intersection grid, and extract the local flow field behavior of the particle material; In the extracted local flow field behavior, the stress transmission chain between the discrete units is constructed, the dynamic breaking and reconstruction process of the stress transmission chain is captured, and the accumulation boundary surface of the granular material at the hopper transfer point is determined; The accumulation boundary surface is divided into a plurality of curvature analysis units, and the actual discharge angle of the granular material is determined through the curvature change rate of the curvature analysis unit; Based on the deviation between the actual discharge angle and the preset target discharge angle, an adjustment matrix driven by the deviation is constructed, the elements of the adjustment matrix include the adjustment amplitude of the local boundary area friction coefficient and the adjustment direction of the local adhesion characteristic of the discrete unit, and the three-dimensional geometric model is iteratively updated through the adjustment matrix until the deviation between the actual discharge angle and the preset target discharge angle is less than a preset deviation threshold, and the optimization modeling result of the belt conveyor is obtained.

[0007] In another aspect, the present application provides a coal mine conveyor operation state intelligent analysis and optimization system for realizing the coal mine conveyor operation state intelligent analysis and optimization method, comprising: A parameter acquisition module is configured to input the geometric structure parameters of the belt conveyor hopper transfer point, the physical property parameters of the granular material, and the environmental humidity parameters in the Bentley software, the physical property parameters include the particle size distribution, density and initial viscosity of the granular material, and a three-dimensional geometric model of the hopper transfer point is constructed based on the geometric structure parameters; A discretization processing module is configured to discretize the physical property parameters of the granular material into a plurality of discrete units to form an initial granular material discrete unit set, and map the discrete unit set into the three-dimensional geometric model; A characteristic updating module is configured to divide the surface of each discrete unit into a plurality of micro-grid areas, and dynamically update the local adhesion characteristic of each discrete unit by the change of the migration path and migration speed of the water film between the micro-grid areas under different environmental humidity parameters; A coefficient distribution module is configured to divide the boundary surface of the hopper transfer point into a plurality of local boundary areas, and dynamically distribute the friction coefficient of each local boundary area according to the contact frequency and contact pressure of each local boundary area with the granular material discrete unit set; An behavior extraction module is configured to simulate the motion process of the granular material discrete unit set in the three-dimensional geometric model based on the updated local adhesion characteristic and the dynamically distributed friction coefficient, capture the intersection characteristics of the discrete unit motion trajectory inside the hopper transfer point by constructing a space-time intersection grid, and extract the local flow field behavior of the granular material; A boundary determination module is configured to determine the accumulation boundary surface of the granular material at the hopper transfer point by constructing the stress transmission chain between the discrete units in the extracted local flow field behavior, and capturing the dynamic breaking and reconstruction process of the stress transmission chain. The angle calculation module is used to divide the accumulation boundary surface into multiple curvature analysis units, and determine the actual discharge angle of the particulate material by the curvature change rate of the curvature analysis units. The iterative optimization module is used to construct a deviation-driven adjustment matrix based on the deviation between the actual discharge angle and the preset target discharge angle. The elements of the adjustment matrix include the adjustment range of the friction coefficient of the local boundary region and the adjustment direction of the local adhesion characteristics of the discrete unit. The three-dimensional geometric model is iteratively updated through the adjustment matrix until the deviation between the actual discharge angle and the preset target discharge angle is less than the preset deviation threshold, thereby obtaining the optimized modeling result of the belt conveyor.

[0008] The technical effects and advantages of the intelligent analysis and optimization method and system for the operating status of coal mine conveyors of this invention are as follows: This invention enhances the predictive ability of particulate material flow behavior, especially when handling viscous materials. It effectively addresses the complex effects of humidity fluctuations, reducing the deviation between the actual discharge angle and model predictions. In real-world industrial scenarios, this invention effectively avoids problems such as uneven material accumulation, poor discharge, or equipment blockage caused by discharge angle deviations. This reduces the risk of increased energy consumption, equipment wear, and operational interruptions due to abnormal material flow, significantly extending equipment lifespan and reducing maintenance costs. Simultaneously, this invention enhances the system's adaptability to humidity changes, maintaining stable material flow performance in extreme environments such as high humidity during rainy seasons or low humidity during droughts, saving companies the additional expenses associated with frequent equipment or process adjustments due to environmental changes. Furthermore, this invention significantly shortens the optimization design cycle for viscous material flow behavior through intelligent means, reducing reliance on expensive experimental verification, lowering R&D costs, and improving the economic efficiency of engineering design. By optimizing material flow and discharge processes, this invention reduces material loss and energy waste caused by inaccurate flow characteristic predictions, enhancing the sustainability of the production process. Attached Figure Description

[0009] Figure 1 This is a logical diagram of a method for intelligent analysis and optimization of the operating status of a coal mine conveyor according to the present invention; Figure 2 This is a schematic diagram of an intelligent analysis and optimization system for the operating status of a coal mine conveyor according to the present invention. Detailed Implementation

[0010] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0011] This application provides a method and system for intelligent analysis and optimization of the operating status of a coal mine conveyor. The system's execution entities include, but are not limited to, belt conveyor design platforms, material flow analysis systems, discrete element simulation platforms, digital twin systems, and industrial equipment optimization platforms, which can be considered as general computing nodes in this application. The data processing platforms include, but are not limited to, at least one material property analysis system, geometric modeling system, and motion simulation system.

[0012] This invention constructs a three-dimensional geometric model of the hopper transfer point of a belt conveyor using Bentley software. It achieves accurate simulation by discretizing particulate materials, considers the influence of ambient humidity on the local adhesion characteristics of materials, dynamically allocates the friction coefficient of the boundary region, captures the material flow field behavior using a spatiotemporal convergence grid, determines the accumulation boundary surface through stress transmission chain analysis, and optimizes the discharge angle as the core indicator. The deviation-driven adjustment matrix is ​​adaptive, and multiple iterations of optimization ensure that the actual discharge angle reaches the preset target. This invention solves the problem of inaccurate prediction of complex behavior of particulate materials in traditional conveyor modeling and provides a scientific basis for the efficient design of belt conveyors.

[0013] Please see Figure 1 This is a logical schematic diagram of a method for intelligent analysis and optimization of the operating status of a coal mine conveyor according to this application. In this embodiment of the invention, a method for intelligent analysis and optimization of the operating status of a coal mine conveyor includes the following steps: Input the geometric structure parameters of the belt conveyor hopper transfer point, the physical property parameters of the granular material, and the environmental humidity parameters into the Bentley software. The physical property parameters include the particle size distribution, density, and initial viscosity of the granular material. Based on the geometric structure parameters, construct a three-dimensional geometric model of the hopper transfer point. The physical property parameters of particulate materials are discretized into multiple discrete units to form an initial set of discrete units of particulate materials, and the set of discrete units is mapped to a three-dimensional geometric model. The surface of each discrete unit is divided into multiple micro-grid regions. Under different environmental humidity parameters, the local adhesion properties of each discrete unit are dynamically updated by the changes in the migration path and migration speed of the moisture film between the micro-grid regions. The boundary surface of the hopper transfer point is divided into multiple local boundary regions. The friction coefficient of each local boundary region is dynamically allocated based on the contact frequency and contact pressure between each local boundary region and the discrete unit set of particulate material. Based on the updated local adhesion characteristics and dynamically allocated friction coefficient, the motion process of discrete unit sets of particulate material is simulated in a three-dimensional geometric model. By constructing a spatiotemporal intersection grid, the intersection characteristics of the motion trajectory of discrete units inside the hopper transfer point are captured, and the local flow field behavior of particulate material is extracted. In the extracted local flow field behavior, by constructing stress transmission chains between discrete units, the dynamic fracture and reconstruction process of stress transmission chains is captured, and the accumulation boundary surface of particulate material at the hopper transfer point is determined. The accumulation boundary surface is divided into multiple curvature analysis units. The actual discharge angle of the particulate material is determined by the curvature change rate of the curvature analysis units. Based on the deviation between the actual discharge angle and the preset target discharge angle, a deviation-driven adjustment matrix is ​​constructed. The elements of the adjustment matrix include the adjustment range of the friction coefficient of the local boundary region and the adjustment direction of the local adhesion characteristics of the discrete unit. The three-dimensional geometric model is iteratively updated by the adjustment matrix until the deviation between the actual discharge angle and the preset target discharge angle is less than the preset deviation threshold, thus obtaining the optimized modeling result of the belt conveyor.

[0014] In this embodiment, a precise 3D model of the belt conveyor hopper transfer point is first created in Bentley software. Complete geometric parameters are required, including belt width (500, 650, 800, 1000, 1200, 1400, 1600, 1800, 2000, 2200, 2400), belt type (STEEL, EP, NN), idler type (trough, forward-inclined, buffer, parallel, buffer bed), idler specifications (89, 108, 133, 159, 194, 219, 10°, 20°, 30°, 35°, 45°), drive type (head single / double drive, middle single / double drive, tail single / double drive, and combinations of the first three), and transmission... The specifications of the moving / redirecting rollers (250, 315, 400, 500, 630, 800, 1000, 1200, 1400, 1600 mm), headstock type (DT2A, DT75, Tangye), tailstock type (DT2A, DT75, Tangye), outrigger type, tensioning device type (screw / worm gear / counterweight / vehicle-type / hydraulic), and key geometric data such as hopper inlet dimensions (width, height), hopper outlet dimensions, hopper wall thickness, hopper inclination angle, radius of curvature of the transition point, and transition point wall material are all crucial. These parameters determine the spatial boundary conditions of the material flow. Simultaneously, the physical properties of the particulate material are input, including particle size distribution (minimum particle size, maximum particle size, particle size probability distribution curve), material density (bulk density and true particle density), and initial viscosity (cohesive force coefficient between materials). These parameters determine the fundamental characteristics of the material itself. In addition, environmental humidity parameters need to be input, including the current relative humidity value, humidity fluctuation range, and humidity change rate. These parameters will affect the flow characteristics of the material. Based on the input geometric parameters, Bentley software uses precise surface modeling technology to construct a three-dimensional geometric model of the hopper transfer point. The model includes complete internal surface details, chamfers, transition areas, and support structures, providing a geometric basis for subsequent simulations.

[0015] The input particulate material physical property parameters are then transformed into discretized computational units. An improved Discrete Element Method (DEM) is employed to discretize the continuous material property parameters into a finite number of computational units. For particle size distribution, multiple spherical or polyhedral discrete units of different diameters are generated based on the probability distribution curve, with the unit size range covering the actual particle size range of the material. For material density, corresponding mass attributes are assigned to each discrete unit to ensure that the overall density distribution conforms to the input parameters. For initial viscosity, contact model parameters between discrete units are set, including the friction coefficient, coefficient of restitution, and adhesion strength. The discretization process employs an adaptive mesh generation algorithm, using a finer discrete mesh in key material flow regions to improve computational accuracy. The resulting initial set of particulate material discrete units typically contains tens of thousands to millions of discrete units, each with independent physical property parameters and state variables. These discrete units are mapped to the feed region of the three-dimensional geometric model according to the initial distribution rules, preparing for subsequent dynamic simulations.

[0016] In this embodiment of the invention, the surface of each discrete unit is divided into multiple micro-grid regions. Under different environmental humidity parameters, the local adhesion characteristics of each discrete unit are dynamically updated by varying the migration path and speed of the moisture film between the micro-grid regions, including: The surface of each discrete unit is divided into multiple micro-grid regions, and an initial moisture film thickness is assigned to each micro-grid region based on the ambient humidity parameters and the initial viscosity of the particulate material. At different time steps, the migration speed and direction of the water film are determined by the migration path of the water film between adjacent microgrid regions. The migration path is driven by the gradient of the water film thickness, and the migration speed is determined by the ratio of the length of the migration path to the time step. The moisture film thickness of each microgrid region is updated based on the migration speed and migration direction. The update of the moisture film thickness is achieved by the cumulative increase or decrease of the moisture film along the migration path. Based on the updated moisture film thickness, the local adhesion force of each microgrid region is adjusted through the interaction between the moisture film thickness and the initial viscosity, wherein the adjustment direction of the local adhesion force is determined by the increasing or decreasing trend of the moisture film thickness. The local adhesion forces of all tiny grid regions on the surface of each discrete unit are integrated to form the local adhesion properties of the discrete unit.

[0017] In this embodiment, the surface of the discrete unit is first finely divided to create a microscale mesh structure. The surface of each discrete unit is divided into tens to hundreds of micro-grid regions. The division method adopts a curvature-adaptive mesh generation algorithm, with higher mesh density in regions where the surface curvature changes significantly to capture fine surface features. The area of ​​each micro-grid region is typically in the range of 0.01-0.1 times the particle size of the material, ensuring sufficient resolution to simulate local moisture distribution. Based on the input environmental humidity parameters and the initial viscosity of the particulate material, the initial moisture film thickness is calculated and assigned to each micro-grid region. The calculation process considers three key factors: the ambient relative humidity value, which determines the available moisture content in the air; the hygroscopic properties of the material, indirectly characterized by the initial viscosity parameter; and the surface micro-geometric features, determined by the curvature and position of the mesh region. These three factors are comprehensively calculated through a preset moisture adsorption model to generate an initial moisture film thickness distribution map, typically expressed as a thickness value at the micrometer level.

[0018] Within each time step of the simulation, the migration process of the moisture film between adjacent microgrid regions is simulated. The migration mechanism is based on the theory of moisture diffusion and is mainly driven by the gradient of the moisture film thickness; that is, moisture always migrates from regions with higher thickness to regions with lower thickness. First, the moisture film thickness gradient between each pair of adjacent grid regions is calculated, forming a thickness gradient field. Based on the gradient field, the main path of moisture migration is determined, i.e., the direction of the steepest gradient. The moisture migration rate is determined by two factors: the magnitude of the thickness gradient (the larger the gradient, the faster the rate); and the microscopic morphological characteristics of the material surface, such as roughness and porosity, which affect the flow resistance of moisture. The migration rate is finally calculated as the ratio of the migration path length to the time step, expressed in micrometers per second, representing the distance the moisture film moves on the surface per unit time.

[0019] The moisture film thickness of each microgrid region is updated based on the calculated migration velocity and direction. The update process adheres to the conservation principle, meaning the amount of water flowing out of one region equals the amount flowing into adjacent regions. For each grid region, the water inflow and outflow rates of all adjacent regions are calculated, and the current region's moisture film thickness is updated based on the net flow rate. The update formula considers the time step, grid region area, and moisture migration velocity to ensure physical plausibility. In cases of varying ambient humidity, moisture exchange with the environment—namely, evaporation and condensation processes—is also considered. The rates of these processes are determined by the difference between the current moisture film thickness and the equilibrium thickness (which is determined by ambient humidity).

[0020] Based on the updated moisture film thickness, the local adhesion force of each microgrid region is adjusted. The adjustment process is based on the mechanism by which moisture affects the adhesion behavior of particulate materials: when the moisture film thickness increases within a specific range, the adhesion force usually increases due to the formation of liquid bridges; when there is too much or too little moisture, the adhesion force may weaken. Using a pre-defined moisture-adhesion force relationship model, the moisture film thickness is mapped to an adhesion force adjustment coefficient. This coefficient is multiplied by the initial viscosity of the material to obtain the adjusted local adhesion force. The direction of adhesion force adjustment is determined by the trend of moisture film thickness variation. The variation pattern of the moisture film thickness in each region over several consecutive time steps is detected, its future trend is predicted, and the direction of adhesion force variation is adjusted accordingly.

[0021] Finally, the local adhesion forces of all tiny grid regions on the surface of each discrete element are integrated to form the overall adhesion properties of that discrete element. The integration process considers the area weight and positional importance of different grid regions; grid regions with larger areas or located in critical positions (such as near contact points) have a greater impact on the overall adhesion properties. The integrated adhesion properties include average adhesion strength, adhesion strength distribution, and directional characteristics. These properties directly affect the interaction behavior of the discrete element with other elements and boundary surfaces in subsequent simulations. This microscale moisture migration simulation accurately captures the complex influence of ambient humidity on the flow behavior of particulate materials, improving the realism of the simulation and the accuracy of predictions.

[0022] In this embodiment of the invention, the friction coefficient of each local boundary region is dynamically allocated based on the contact frequency and contact pressure between each local boundary region and the discrete unit set of particulate material, including: The boundary surface of the hopper transfer point is divided into multiple local boundary regions, and the initial friction coefficient is assigned according to the surface roughness of each local boundary region. Within a unit of time, the number of contacts between each local boundary region and the discrete unit set of particulate material is counted to form a contact frequency level, wherein the contact frequency level is determined by the relative magnitude of the number of contacts. On a unit area, the force distribution of each local boundary region is calculated to form a contact pressure level, where the contact pressure level is determined by the peak value of the force distribution; Based on the contact frequency level and contact pressure level, the friction coefficient of each local boundary region is adjusted. When the contact frequency level is higher than the preset frequency threshold, the friction coefficient is increased, and when the contact pressure level is higher than the preset pressure threshold, the friction coefficient is decreased. The adjusted friction coefficient is used to form the dynamic friction coefficient for each local boundary region.

[0023] In this embodiment, the boundary surface of the hopper transfer point is first rationally divided to create local boundary region meshes. The division process employs an adaptive mesh generation algorithm, adjusting the mesh density based on surface geometry (such as curvature and angle variations) and the expected material flow path. More refined meshes are used in critical material flow areas (such as transfer bends and near the discharge port). The size of each local boundary region is typically controlled within 5-10 times the average particle size of the material to ensure accurate capture of local interaction characteristics. For each divided local boundary region, an initial friction coefficient is assigned based on its surface roughness characteristics. Surface roughness can be extracted from the material properties of the Bentley model or specified by user input. The initial friction coefficient is assigned using an empirical mapping relationship, mapping surface roughness values ​​(usually represented by Ra values) to appropriate friction coefficients, typically ranging from 0.1 to 0.8, with rougher surfaces assigned higher friction coefficients.

[0024] During the simulation, the interaction between each local boundary region and the discrete unit set of particulate material is continuously monitored. Within each unit time window (typically 0.1-1 seconds of the simulation time), the contact events between each local boundary region and the discrete unit set of particulate material are precisely counted. Contact statistics include three key parameters: contact count, recording the total number of contact events occurring per unit time; contact duration, recording the duration of each contact; and contact area, recording the surface area involved in each contact. Based on these raw statistical data, a contact frequency index is calculated, which is the weighted number of contacts per unit time, with the weights determined by the contact duration and contact area. The calculated contact frequency values ​​are divided into multiple levels (typically 5-10 levels) according to their relative magnitude, forming a contact frequency level distribution map, which visually displays the contact hotspot areas on the boundary surface.

[0025] Simultaneously, the force distribution of each local boundary region is calculated. The calculation process is based on the contact force model of the discrete element method, accurately tracking the normal and tangential forces generated when each discrete element contacts the boundary surface. All contact forces are accumulated per unit area to form a force density distribution map, with units of N / m². The force distribution analysis focuses on three key indicators: peak pressure, the maximum force density value within the region; average pressure, the average force density within the region; and pressure gradient, the rate of change of force density in space. Based on the relative magnitude of the peak pressure, each local boundary region is divided into multiple contact pressure levels (typically 5-10 levels), forming a pressure level distribution map that visually displays the high-pressure and low-pressure areas on the boundary surface.

[0026] Based on the acquired contact frequency and pressure levels, the friction coefficient of each local boundary region is dynamically adjusted. The adjustment strategy follows two basic principles: when the contact frequency level is higher than a preset frequency threshold (usually set to a medium frequency level), the friction coefficient of the corresponding region is increased, reflecting the surface wear and material deposition effects that may occur in high-frequency contact areas; when the contact pressure level is higher than a preset pressure threshold (usually set to a medium pressure level), the friction coefficient of the corresponding region is decreased, reflecting the surface smoothing and material compaction effects that may occur in high-pressure areas. The adjustment range of the friction coefficient is proportional to the degree to which it exceeds the threshold, and the adjustment formula includes a non-linear factor to ensure a reasonable upper limit for the adjustment range in extreme cases. For regions that simultaneously meet the conditions of high frequency and high pressure, the influence of the two factors is balanced according to preset weighting coefficients, with the pressure factor typically having a slightly higher weight than the frequency factor.

[0027] Through the above adjustment process, a dynamic friction coefficient is generated for each local boundary region. The dynamic friction coefficient reflects the true characteristics of the interaction between the boundary surface and the material under actual working conditions, and is more accurate than the statically assigned friction coefficient. The dynamic friction coefficient is saved as a time-varying parameter and updated periodically to reflect the gradual change in the boundary surface characteristics. This dynamic friction coefficient allocation mechanism enables the model to accurately simulate the wear and material accumulation effects on the hopper transfer point surface during long-term operation, improving the simulation's realism and prediction accuracy.

[0028] In this embodiment of the invention, by constructing a spatiotemporal intersection grid, the intersection characteristics of discrete unit motion trajectories within the hopper transfer point are captured, and the local flow field behavior of particulate materials is extracted, including: The interior of the hopper transfer point is divided into a spatiotemporal intersection grid, which consists of multiple spatiotemporal units. Each spatiotemporal unit corresponds to a local spatial region and a time segment inside the hopper transfer point. Within each spatiotemporal unit, the motion trajectory of the discrete unit is decomposed into a motion direction vector and a motion velocity vector. The intersection complexity of the spatiotemporal unit is determined by the angular distribution of the motion direction vector and the gradient distribution of the motion velocity vector. The intersection complexity is jointly determined by the concentration of the angular distribution and the fluctuation amplitude of the gradient distribution. In the spatiotemporal intersection grid, spatiotemporal units with intersection complexity exceeding a preset complexity threshold are identified to form high intersection regions; Within the high convergence region, the local flow field behavior of particulate materials is extracted by the spatiotemporal evolution of the motion direction vector and the motion velocity vector. The local flow field behavior is characterized by the velocity distribution and shear stress distribution in the high convergence region.

[0029] In this embodiment, an innovative spatiotemporal convergence grid structure is first established to achieve high-precision capture of the flow characteristics of particulate materials. The internal space of the hopper transfer point is divided into regular or irregular three-dimensional grids. The grid division method is adaptively adjusted according to the geometric characteristics, and a finer grid division is used in areas where the expected material flow is complex (such as transfer bends and contraction areas). Simultaneously, the simulation time axis is divided into multiple time segments, with a time resolution typically in the range of 0.01-0.1 seconds, ensuring the capture of transient characteristics of high-speed flow. By combining the spatial grid with the time segments, a four-dimensional spatiotemporal convergence grid is formed, where each spatiotemporal unit represents the state of a specific location within a specific time window. This spatiotemporal grid structure can simultaneously analyze the spatial distribution and temporal evolution characteristics of material flow, providing a basic framework for subsequent analysis of complex flow behavior.

[0030] Within each spatiotemporal cell, the motion characteristics of all discrete cells passing through that cell are precisely recorded and analyzed. The analysis first decomposes the motion trajectory of each discrete cell into two key vectors: a motion direction vector, representing the spatial direction of the discrete cell's movement, typically represented by a unit vector; and a motion velocity vector, representing the magnitude and direction of the discrete cell's velocity, containing complete velocity information. For each spatiotemporal cell, vector information from all discrete cells passing through that cell is collected, forming a set of direction vectors and a set of velocity vectors. Based on these vector sets, two key indicators are calculated: the angular distribution of the direction vectors, which is generated by calculating the angle between any two direction vectors to form an angular frequency distribution histogram, reflecting the uniformity or divergence of the flow direction; and the gradient distribution of the velocity vectors, which is generated by calculating the velocity differences between adjacent discrete cells to form a velocity gradient distribution map, reflecting the uniformity or non-uniformity of the flow velocity.

[0031] The convergence complexity index of spatiotemporal units is calculated using angular distribution and gradient distribution. Convergence complexity is a comprehensive indicator characterizing the degree of disorder in material flow, consisting of two sub-indicators: the concentration of the angular distribution, obtained by calculating the entropy or standard deviation of the angular distribution; the more dispersed the distribution (the higher the entropy or the larger the standard deviation), the more disordered the flow direction and the lower the concentration; and the fluctuation amplitude of the gradient distribution, obtained by calculating the range or coefficient of variation of the velocity gradient; the larger the fluctuation, the more drastic the change in flow velocity. These two sub-indicators are weighted and summed to generate the final convergence complexity value, which is usually normalized to a range of 0-1, with higher values ​​indicating more complex flow. The calculated convergence complexity is visualized as a heatmap, intuitively presenting the distribution of flow complexity within the hopper transfer point.

[0032] Within the constructed spatiotemporal convergence grid, spatiotemporal units with convergence complexity exceeding a preset threshold are identified; these units constitute high convergence regions. Threshold settings are typically based on statistical analysis, such as taking the upper quartile of the complexity distribution or the mean plus a standard deviation. High convergence regions represent spatiotemporal areas with particularly complex material flows, potentially corresponding to critical inflection points, bottlenecks, or unstable regions. These high convergence regions are labeled and classified according to their spatiotemporal distribution characteristics into different types, such as persistent high convergence regions (existing in the same spatial location for a long time), mobile high convergence regions (moving and evolving in space), and pulsed high convergence regions (appearing for a short time and then disappearing). Different types of high convergence regions may indicate different flow problems, requiring targeted analysis.

[0033] Within the identified high-convergence regions, more refined analysis is used to extract the local flow field behavior of particulate materials. The analysis focuses on two key indicators: velocity distribution, including the spatial distribution and temporal variation of velocity magnitude, as well as the spatial variation and temporal evolution of velocity direction; and shear stress distribution, calculated by measuring the shear rate through the relative motion between adjacent discrete units, and then combining this with material properties to calculate shear stress, forming a stress distribution map. By analyzing the spatiotemporal evolution characteristics of these indicators, various typical local flow field behaviors are identified, such as flow stagnation, flow bifurcation, flow swirling, and flow oscillation. The identification and classification of these local flow field behaviors provide a direct basis for subsequent optimization. For different types of flow field problems, corresponding improvement suggestions can be generated, such as adjusting geometry, changing surface properties, or optimizing operating parameters.

[0034] In this embodiment of the invention, by constructing a stress transmission chain between discrete units and capturing the dynamic fracture and reconstruction process of the stress transmission chain, the accumulation boundary surface of particulate material at the hopper transfer point is determined, including: In a set of discrete units of particulate material, a stress transmission chain is constructed by identifying the contact points between each discrete unit and its adjacent discrete units. The stress transmission chain consists of a sequence of contact force directions and magnitudes at the contact points. At different time steps, the local stability of the stress transmission chain is determined by the deflection angle of the contact force direction and the fluctuation range of the contact force magnitude at the contact point. The local stability is determined by the weighted sum of the deflection angle and the fluctuation range. In the stress transmission chain, when the deflection angle is greater than a preset angle threshold or the fluctuation amplitude is greater than a preset amplitude threshold, the dynamic break point is identified, and the reconstruction path of the stress transmission chain is determined by the contact point formation between the discrete unit after the dynamic break point and the new adjacent discrete unit. By analyzing the dynamic fracture and reconstruction paths of the stress transmission chain, the packing boundary surface of particulate materials is determined, where the packing boundary surface is determined by the spatial distribution of the fracture points and reconstruction paths of the stress transmission chain.

[0035] In this embodiment, a complex stress transmission network is first identified and constructed within the set of discrete units of the particulate material. The contact state between each discrete unit and its surrounding discrete units is precisely detected, including the spatial location of the contact points, the direction of the contact normal, and the contact area. For each effective contact point, the magnitude and direction of the contact force are calculated. The contact force calculation is based on a selected contact force model (such as the Hertzian contact model, the JKR adhesion model, etc.), considering normal elastic deformation, tangential friction, and local adhesion effects. Among all contact points, those bearing the main load are identified, i.e., points where the contact force exceeds a preset threshold (typically 1.5-2 times the average contact force). These contact points constitute the main stress transmission paths. By connecting these key contact points, a stress transmission chain network is constructed. Each stress transmission chain consists of an ordered sequence of contact force directions and magnitudes, visually representing the propagation path and distribution pattern of stress in the particulate material.

[0036] In each time step of the simulation, the dynamic evolution of the stress transmission chain is monitored. Two key indicators are monitored: the deflection angle of the contact force direction, representing the degree of change in the contact force direction within adjacent time steps, usually expressed as an angle value (degrees); and the fluctuation range of the contact force magnitude, representing the proportion of change in the contact force magnitude within adjacent time steps, usually expressed as a percentage. These two indicators reflect the local stability of the stress transmission chain; the smaller the indicator value, the more stable the chain. Using a pre-defined weighted formula, these two indicators are combined into a local stability index, with the value typically normalized to 0-1; a higher value indicates greater stability. The spatial distribution of the stability index forms a stress stability map, visually displaying the stable and unstable regions of the stress structure in the particulate material.

[0037] In a stress transmission chain network, vulnerable points that may fracture are accurately identified. The identification criteria are based on two conditions: the deflection angle of the contact force direction at the contact point exceeds a preset angle threshold (typically set at 15-30 degrees), indicating a significant change in the contact state; and the fluctuation range of the contact force at the contact point exceeds a preset amplitude threshold (typically set at 20-40%), indicating unstable load-bearing capacity. Contact points meeting either condition are marked as potential fracture points. These potential fracture points are continuously monitored, and when the fracture conditions are met for multiple consecutive time steps, they are confirmed as dynamic fracture points. For each confirmed fracture point, its spatial location, occurrence time, pre-fracture stress state, and surrounding material distribution are recorded to provide detailed data for subsequent analysis.

[0038] After the stress transmission chain breaks, the process of contact formation between the discrete unit after the break point and surrounding new discrete units is tracked; this is the stress transmission chain reconstruction process. The tracking process focuses on three key parameters: the formation time of the new contact point, representing the time interval from the occurrence of the break to the formation of the new contact; the spatial distribution of the new contact points, representing the spatial positional relationship of the newly formed contact points relative to the original break point; and the magnitude and direction of the new contact force, representing the stress transmission characteristics after reconstruction. These parameters are used to evaluate the stability and efficiency of the reconstruction path. A highly stable reconstruction path indicates that the material can quickly form a new supporting structure, while a highly efficient reconstruction path indicates that stress can effectively bypass the break region and continue to be transmitted. The spatiotemporal evolution characteristics of the reconstruction path reflect the self-organizing ability and structural adaptability of the material during dynamic flow.

[0039] By comprehensively analyzing the distribution of breakpoints and reconstruction path characteristics of the stress transmission chain, the accumulation boundary surface of particulate material at the hopper transfer point is determined. The determination process is based on two key information sources: the spatial distribution of breakpoints in the stress transmission chain, which are typically concentrated at the boundary between the accumulated mass and the flow region, marking the boundary of the supporting structure; and the spatial distribution of reconstruction paths, which describe the patterns of material reorganization to form a stable structure. A spatial clustering algorithm is used to classify breakpoints and reconstruction paths with similar characteristics, and then a three-dimensional surface fitting algorithm is used to generate a continuous accumulation boundary surface. This surface accurately describes the accumulation morphology of the material within the hopper transfer point and is a crucial basis for determining the discharge angle and optimizing the design.

[0040] In this embodiment of the invention, the actual discharge angle of the particulate material is determined by the rate of change of curvature of the curvature analysis unit, including: The stacked boundary surface is divided into multiple curvature analysis units, and initial curvature values ​​are assigned according to the local geometry of each curvature analysis unit. At different time steps, the curvature change rate is determined by the difference between the curvature values ​​of the curvature analysis unit and the time step. The curvature change rate is determined by the ratio of the difference to the time step. In the stacked boundary surface, curvature analysis units with curvature change rates higher than a preset change rate threshold are identified, forming high curvature change regions; Within the region of high curvature variation, the local unloading angle is determined by the angle distribution between the tangent direction and the horizontal plane, where the local unloading angle is determined by the weighted average of the angle distribution. The local discharge angles of all high curvature variation regions in the stacking boundary surface are integrated to form the actual discharge angle of the particulate material.

[0041] In this embodiment, the determined accumulation boundary surface is first finely divided to create a regional mesh suitable for curvature analysis. An adaptive surface subdivision algorithm is used to adjust the mesh density based on the local geometric features of the surface (such as curvature changes and surface undulations), employing a finer mesh in areas with significant geometric changes. The size of each curvature analysis cell is typically controlled within the range of 3-8 times the average particle size of the material, ensuring sufficient spatial resolution to capture surface features while smoothing out random fluctuations in local particle arrangement. For each subdivided curvature analysis cell, an initial curvature value is calculated and assigned based on its local geometry. The curvature calculation uses a differential geometry method, focusing on two main indices: mean curvature, representing the average degree of curvature of the surface at that point; and Gaussian curvature, representing the inherent curvature characteristics of the surface at that point. These two curvature values ​​together describe the local shape characteristics of the surface, providing a benchmark for subsequent variation analysis.

[0042] In the continuous time steps of the simulation, the curvature value change of each curvature analysis unit is monitored. For each pair of adjacent time steps (t and t+Δt), the difference in curvature values ​​ΔK = K(t+Δt) - K(t) is calculated, which reflects the degree of change in the surface shape within this time interval. Then, the rate of change of curvature, i.e., the ratio of the curvature difference to the time step, is calculated in units of 1 / s, representing the rate of change of curvature per unit time. The calculation of the rate of change of curvature is applied to both the mean curvature and Gaussian curvature, forming two independent rate of change indices. The spatial distribution of these rate of change indices forms a rate of change of curvature map, visually displaying the regions on the packing boundary surface where shape changes are most active. High rate of change regions typically correspond to critical transition regions of material flow, such as the boundary from rest to motion, or the critical zone from stable packing to unstable slip.

[0043] On the packing boundary surface, curvature analysis units with curvature change rates exceeding a preset threshold are identified; these units constitute high curvature variation regions. Threshold settings are typically based on statistical analysis, such as taking the upper quartile of the rate of change distribution or the mean plus two standard deviations. High curvature variation regions represent the most unstable or active parts of the packing boundary surface and are often key areas for determining the discharge angle. These high-curvature variation regions are classified and clustered, grouping spatially adjacent high-curvature variation units into the same group to form continuous high-curvature variation region blocks. The shape, size, and location characteristics of each region block are recorded to provide spatial reference for subsequent discharge angle analysis.

[0044] Within each identified high curvature variation region, the local discharge angle is determined by analyzing the tangent direction. For each curvature analysis unit within the region, the angle between its surface tangent direction and the horizontal plane is calculated; this angle represents the degree of inclination of the material surface at that point. All tangent angle values ​​are collected across the entire high curvature variation region, forming an angle distribution histogram. Based on the angle distribution characteristics, representative local discharge angle values ​​are calculated. A weighted average is used for the calculation, with weighting factors including: the magnitude of the curvature change rate (higher rates have greater weight); the area of ​​the curvature analysis unit (larger areas have greater weight); and the locational importance of the curvature analysis unit, such as units closer to the discharge area having higher weight. This weighted average method yields local angle values ​​that more accurately represent the actual discharge characteristics.

[0045] Finally, the local discharge angles of all high curvature variation regions on the stacking boundary surface are integrated to form the overall actual discharge angle of the particulate material. The integration process considers the spatial distribution and relative importance of each region, typically employing a segmented weighted average method. Higher weights are assigned to high curvature variation regions near the discharge area, as these regions directly affect the material's discharge behavior; lower weights are assigned to high curvature variation regions far from the discharge area, as these regions may reflect internal flow characteristics more than discharge characteristics. This spatial weighting method calculates the final actual discharge angle value, which comprehensively considers the complex flow and stacking behavior of the material within the hopper transfer point, and more accurately reflects the discharge characteristics under actual working conditions than simple static angle measurements.

[0046] In this embodiment of the invention, a deviation-driven adjustment matrix is ​​constructed. The elements of the adjustment matrix include the adjustment magnitude of the friction coefficient in the local boundary region and the adjustment direction of the local adhesion characteristics of the discrete unit, including: Inside the hopper transfer point, a deviation distribution map is constructed by the deviation between the actual discharge angle and the preset target discharge angle. The deviation distribution map consists of the deviation values ​​of multiple local areas. In the deviation distribution map, the adjustment range of the friction coefficient is determined by the local correlation between the deviation value and the friction coefficient of the local boundary region, where the adjustment range is determined by the strength of the correlation. In the deviation distribution diagram, the adjustment direction of the local adhesion characteristics is determined by the local correlation between the deviation value and the local adhesion characteristics of the discrete unit, where the adjustment direction is determined by the positive or negative sign of the correlation. By adjusting the friction coefficient and the direction of local adhesion characteristics, a deviation-driven adjustment matrix is ​​constructed, where the rows of the adjustment matrix correspond to the local boundary region of the hopper transfer point, and the columns correspond to the local adhesion characteristics of the discrete unit set of particulate material. By adjusting the matrix, the update strategy for the three-dimensional geometric model is determined.

[0047] In this embodiment, a detailed deviation distribution map is first constructed by comparing the actual discharge angle with the preset target discharge angle. The stacking boundary surface is divided into multiple local analysis regions, each corresponding to a specific location on the surface. For each local region, the deviation value between the local actual discharge angle and the preset target discharge angle is calculated. The deviation value can be an absolute difference (degrees) or a relative difference (percentage). The spatial distribution of these local deviation values ​​forms a deviation distribution map, which visually displays the discharge angle deviation at different locations on the stacking boundary surface. The deviation distribution map is visualized as a heatmap, with colors ranging from blue to red representing negative deviation (actual angle less than the target angle) to positive deviation (actual angle greater than the target angle). This detailed deviation distribution analysis can identify areas that require focused adjustment, rather than simply adjusting based on the overall average deviation.

[0048] In the constructed deviation distribution map, the correlation between the deviation value of each local region and the friction coefficient of the corresponding local boundary region is analyzed. The analysis employs a local correlation analysis method. For each local region, all data points within a predetermined radius are collected, including the deviation value and friction coefficient value within that radius. Then, the Pearson correlation coefficient or Spearman's rank correlation coefficient between these local data points is calculated to obtain a local correlation index. This index ranges from -1 to 1, with the absolute value representing the correlation strength and the sign indicating the correlation direction (positive or negative). Based on the calculated local correlation strength, the adjustment range of the friction coefficient is determined. The adjustment range is proportional to the correlation strength; the stronger the correlation, the larger the adjustment range, ensuring that adjustment resources are concentrated on the most significantly affected parameters. The adjustment range is typically standardized using a nonlinear mapping function (such as the sigmoid function) to keep it within a reasonable range.

[0049] Similarly, the local correlation between the deviation value and the local adhesion characteristics of the discrete unit is analyzed. The analysis method is similar to the friction coefficient correlation analysis, but focuses on adhesion characteristic parameters, such as moisture film thickness and local adhesion force. Through local correlation analysis, the correlation coefficient between the deviation value and the adhesion characteristics is obtained. The sign of this coefficient determines the direction of adjustment of the local adhesion characteristics. When the correlation coefficient is positive, it indicates that increasing the adhesion characteristics may help reduce the deviation, and the adjustment direction is set to positive (increase); when the correlation coefficient is negative, it indicates that decreasing the adhesion characteristics may help reduce the deviation, and the adjustment direction is set to negative (decrease). The determination of the adjustment direction takes into account the physical mechanism of material flow to ensure that the adjustment conforms to physical laws and does not lead to unreasonable parameter combinations.

[0050] Based on the determined adjustment range of the friction coefficient and the adjustment direction of local adhesion characteristics, a complete deviation-driven adjustment matrix is ​​constructed. This matrix is ​​a two-dimensional data structure, where rows correspond to local boundary regions at the hopper transfer point (typically dozens to hundreds of regions), and columns correspond to the local adhesion characteristic categories of the discrete unit set of particulate materials (such as adhesion characteristics under different locations and humidity conditions). Each element aij in the matrix represents the joint adjustment parameter of the friction coefficient and the j-th type of local adhesion characteristic for the i-th local boundary region. The element value consists of two parts: the adjustment range of the friction coefficient and the adjustment direction of the adhesion characteristic. These two values ​​are synthesized into a single adjustment parameter through a preset combination rule (such as weighted sum or product). The adjustment matrix is ​​constructed using sparse matrix technology, storing only elements with significant correlations to improve computational efficiency.

[0051] The specific update strategy for the 3D geometric model is determined by constructing an adjustment matrix. The update strategy includes three main aspects: parameter adjustment, which modifies the corresponding friction coefficient and adhesion characteristic parameters based on the element values ​​in the adjustment matrix, with the adjustment magnitude proportional to the matrix element values; geometric fine-tuning, which may suggest fine-tuning local geometry for areas where the adjustment effect is unsatisfactory, such as changing the wall angle, adding flow guiding structures, or modifying the surface morphology; and structural optimization, which includes optimizing parameters such as bandwidth, idler spacing, and support leg arrangement to improve material flow characteristics. The update strategy is then translated into specific execution instructions and applied to the 3D geometric model to generate a new model version. The updated model will enter the next round of simulation and evaluation loops until the deviation between the actual discharge angle and the preset target discharge angle meets the preset convergence condition.

[0052] The optimized modeling results of the belt conveyor obtained by this method can not only accurately predict the flow behavior of particulate materials at the hopper transfer point, but also provide direct guidance for the design and improvement of belt conveyors. The optimization results can be applied to adjusting the belt width selection, optimizing the idler spacing, improving the support leg design, optimizing the head and tail frame configuration, and selecting the tensioning device, thereby improving the operating efficiency of the belt conveyor, reducing material blockage and overflow problems, and extending the equipment's service life. Compared with traditional empirical design methods, this optimization method based on accurate physical simulation can better adapt to different working conditions and material characteristics, providing a more reliable and efficient belt conveyor design solution.

[0053] In this embodiment of the invention, in a spatiotemporal intersection grid, identifying spatiotemporal units with intersection complexity higher than a preset complexity threshold to form a high intersection region includes: Within each spatiotemporal cell of the spatiotemporal intersection grid, an intersection complexity distribution map is constructed by the concentration of the angle distribution of the motion direction vector and the fluctuation amplitude of the gradient distribution of the motion velocity vector. The intersection complexity distribution map is composed of a weighted sum of the concentration of the angle distribution and the fluctuation amplitude of the gradient distribution. In the intersection complexity distribution map, the spatiotemporal units corresponding to local peaks are determined by identifying spatiotemporal units with concentrations higher than a preset concentration threshold or fluctuation amplitudes higher than a preset fluctuation amplitude threshold. By integrating the spatiotemporal units corresponding to local peaks, a high convergence region is formed in the spatiotemporal convergence grid.

[0054] In this embodiment, a detailed vector analysis is first performed within each spatiotemporal cell of the spatiotemporal intersection grid to construct an accurate intersection complexity distribution map. For each spatiotemporal cell, motion information from all discrete cells passing through that cell is collected, constructing two key statistical distributions: the angle distribution of motion direction vectors and the gradient distribution of motion velocity vectors. For the angle distribution, the angle between the motion direction vectors of any two discrete cells within the cell is calculated, generating an angle frequency distribution histogram. Then, the concentration index of the angle distribution is calculated using information entropy or distribution standard deviation. The lower the concentration (the higher the entropy value or the larger the standard deviation), the more dispersed the directions are, and the worse the consistency of the flow direction. For the gradient distribution, the velocity difference between adjacent discrete cells within the cell is calculated, generating a velocity gradient frequency distribution histogram. Then, the fluctuation amplitude of the gradient distribution is calculated using the coefficient of variation or kurtosis index. The larger the fluctuation amplitude, the more drastic the velocity change and the worse the flow uniformity.

[0055] Using a pre-defined weighting formula, the concentration index of the angled distribution and the fluctuation amplitude index of the gradient distribution are combined into a single convergence complexity value. The weighting coefficients are optimized based on different application scenarios and material characteristics, typically assigning slightly higher weights to the index that better reflects the key characteristics of material flow, while balancing the contributions of both. The calculated convergence complexity value is standardized and mapped to a range of 0-1, with higher values ​​indicating more complex flows. The convergence complexity values ​​of all spatiotemporal units are organized into a four-dimensional data structure (three spatial dimensions plus a time dimension), forming a complete convergence complexity distribution map. This distribution map is visually displayed using multi-dimensional visualization techniques (such as 3D slice plots and time-series animations), enabling engineers to fully understand the complexity distribution patterns of material flows.

[0056] In the constructed intersection complexity distribution map, key spatiotemporal units with significant complexity are identified using a dual-threshold method. Two independent thresholds are set: a preset concentration threshold, typically set as the upper quartile or mean of the concentration distribution plus one standard deviation; and a preset fluctuation amplitude threshold, typically set as the upper quartile or mean of the fluctuation amplitude distribution plus one standard deviation. A spatiotemporal unit is marked as a potentially high-complexity unit if it meets either of the following conditions: the concentration of the angular distribution is higher than the preset concentration threshold, indicating high directional disorder; or the fluctuation amplitude of the velocity gradient is higher than the preset fluctuation amplitude threshold, indicating drastic velocity changes. Among the marked potentially high-complexity units, local peak units are further selected, i.e., units whose complexity values ​​are higher than all their neighboring units. These local peak units represent the spatial maxima of intersection complexity, indicating the locations where flow complexity is most concentrated.

[0057] Spatial clustering algorithms are used to integrate identified local peak units, forming continuous high-convergence regions. The clustering process employs density-based spatial clustering algorithms (such as DBSCAN) to group spatially similar peak units into the same cluster. Clustering parameters (such as minimum distance threshold and minimum number of points) are adaptively set according to material characteristics and model resolution, ensuring that the generated high-convergence regions are both physically meaningful and not excessively dispersed or merged. For each generated high-convergence region, its geometric features (such as volume, surface area, and centroid location) and statistical features (such as average complexity and complexity standard deviation) are calculated, providing comprehensive region description information. Simultaneously, the evolutionary characteristics of high-convergence regions over time are analyzed, including region formation time, duration, movement trajectory, and splitting / merging events, revealing the dynamic development process of material flow complexity. These spatiotemporally integrated high-convergence regions provide crucial reference for subsequent flow field analysis and optimization design.

[0058] In this embodiment of the invention, the reconstruction path of the stress transmission chain is determined by the contact point formation between the discrete element after the dynamic fracture point and the new adjacent discrete element, including: Around the dynamic fracture point, the contact point between the discrete element after the dynamic fracture point and the new adjacent discrete element is determined by constructing a local neighborhood. The local neighborhood consists of multiple discrete elements around the dynamic fracture point. Within a local neighborhood, the probability of forming a new contact point is determined by the stability of the contact force direction at the new contact point, where the probability of formation is determined by the reciprocal of the deflection angle of the contact force direction. The reconstructed path of the stress transmission chain is constructed by the probability of the formation of new contact points, wherein the reconstructed path consists of a sequence of new contact points and a dynamic adjustment process of the contact force direction.

[0059] In this embodiment, a detailed local analysis region is first constructed around the confirmed dynamic fracture point. Centered on the dynamic fracture point, an appropriate search radius (typically 3-5 times the average particle size of the material) is set to determine the local neighborhood. Within this neighborhood, the motion trajectory of the discrete unit after the fracture point (i.e., the unit connected to that point before fracture) and the distribution of all potential contact units in the surrounding area are accurately tracked. The tracking process employs high temporal resolution (typically an order of magnitude smaller than the conventional simulation time step) to ensure the capture of rapidly changing contact dynamics. The position, velocity, acceleration, and rotation state of the discrete unit after the fracture point are recorded, while the corresponding parameters of all other discrete units in the neighborhood are monitored, forming a complete local dynamic dataset. Based on this dynamic data, it is predicted which new neighboring units the discrete unit after the fracture point may form contact with, considering factors such as the relative position, relative velocity, and geometry of the units.

[0060] For each predicted new contact point, its stability and persistence are assessed. The assessment focuses on the stability of the contact force direction, i.e., the degree of change in the direction of the newly formed contact force over consecutive time steps. The deflection angle of the contact force direction between adjacent time steps is calculated; a smaller angle indicates greater stability. Based on statistical analysis of the deflection angles over multiple consecutive time steps, the average deflection angle and the standard deviation of the deflection angle are calculated. These two indicators together reflect the overall stability of the contact force direction. A physics-based probabilistic model is used to map the stability of the contact force direction to the probability of contact point formation. The mapping function is based on the reciprocal relationship of the deflection angle; that is, the smaller the deflection angle (the more stable the direction), the higher the probability of formation. This probabilistic representation considers the randomness and uncertainty of the material contact process, and is more consistent with the actual physical process than a simple deterministic judgment.

[0061] Based on the calculated contact point formation probability, possible reconfiguration paths for the stress transfer chain are constructed. The construction process employs a path-first search algorithm, starting from the fracture point and expanding outwards along the sequence of new contact points with the highest formation probability. Each expansion considers multiple possible branch paths, forming a tree-like path network. For each potential path, its overall probability (the product of the probabilities of each node) and physical plausibility indicators (such as force balance and energy efficiency) are calculated. Through comprehensive scoring, the most probable reconfiguration path—that is, the path with a high formation probability and good physical plausibility—is identified. In cases where multiple competing paths exist, several possible reconfiguration paths may be retained, reflecting the diversity and uncertainty of the material reconfiguration process.

[0062] A complete description of the reconfiguration path comprises two key parts: a new sequence of contact points, describing how stress is transmitted through the newly formed contact network, with each contact point including its location, contact area, and contact type (point contact, surface contact, etc.); and a dynamic adjustment process of the contact force direction, describing how the contact force adjusts over time to adapt to the new structural equilibrium, including time evolution data of force magnitude and direction. The dynamic development of the reconfiguration path is visualized through animation or time-series charts, intuitively demonstrating how the stress transmission chain rapidly self-repairs and reassembles after fracture. This detailed reconfiguration path analysis reveals the self-organizing ability and structural adaptability of particulate materials during dynamic flow, providing a deeper perspective for understanding the stability mechanisms of material flow.

[0063] During the construction of the 3D geometric model, the system simultaneously generates 3D models of the headframe, tailframe, upper and lower rollers, support legs, and intermediate frame within the selected bandwidth. The headframe, tailframe, and associated rollers and support legs are automatically matched according to different power and tension forces. The upper rollers are arranged at 1200mm intervals except for the convex arc section and the material drop point, while the lower rollers are arranged at 3000mm intervals except for the convex arc section, ensuring that the spacing between the upper rollers does not exceed 1200mm and the lower rollers does not exceed 3000mm. The support legs are arranged at intervals not exceeding 3000mm. The height of the headframe, tailframe, and tensioning device is adjusted in real time based on their actual heights and then assembled into their positions. A guide chute is generated based on the input material drop point location. This automated construction process ensures the accuracy and practicality of the model, providing a reliable geometric basis for subsequent optimization analysis.

[0064] The above describes a method for intelligent analysis and optimization of the operating status of a coal mine conveyor according to an embodiment of this application. The following describes a system for intelligent analysis and optimization of the operating status of a coal mine conveyor according to an embodiment of this application. Please refer to [link / reference]. Figure 2 One embodiment of the intelligent analysis and optimization system for the operating status of a coal mine conveyor in this application includes: The parameter acquisition module is used to input the geometric structure parameters of the belt conveyor hopper transfer point, the physical property parameters of the granular material, and the environmental humidity parameters into the Bentley software. The physical property parameters include the particle size distribution, density, and initial viscosity of the granular material, and a three-dimensional geometric model of the hopper transfer point is constructed based on the geometric structure parameters. The discretization module is used to discretize the physical property parameters of particulate materials into multiple discrete units to form an initial set of discrete units of particulate materials, and to map the set of discrete units to the three-dimensional geometric model. The feature update module is used to divide the surface of each discrete unit into multiple micro-grid regions. Under different environmental humidity parameters, the local adhesion characteristics of each discrete unit are dynamically updated by the changes in the migration path and migration speed of the moisture film between the micro-grid regions. The coefficient allocation module is used to divide the boundary surface of the hopper transfer point into multiple local boundary regions, and dynamically allocate the friction coefficient of each local boundary region according to the contact frequency and contact pressure between each local boundary region and the discrete unit set of particulate material. The behavior extraction module is used to simulate the motion process of discrete unit sets of particulate material in the three-dimensional geometric model based on the updated local adhesion characteristics and dynamically allocated friction coefficient. By constructing a spatiotemporal intersection grid, it captures the intersection characteristics of the motion trajectory of discrete units inside the hopper transfer point and extracts the local flow field behavior of particulate material. The boundary determination module is used to determine the accumulation boundary surface of particulate material at the hopper transfer point by constructing stress transmission chains between discrete units in the extracted local flow field behavior, capturing the dynamic fracture and reconstruction process of the stress transmission chains; The angle calculation module is used to divide the accumulation boundary surface into multiple curvature analysis units, and determine the actual discharge angle of the particulate material by the curvature change rate of the curvature analysis units. The iterative optimization module is used to construct a deviation-driven adjustment matrix based on the deviation between the actual discharge angle and the preset target discharge angle. The elements of the adjustment matrix include the adjustment range of the friction coefficient of the local boundary region and the adjustment direction of the local adhesion characteristics of the discrete unit. The three-dimensional geometric model is iteratively updated through the adjustment matrix until the deviation between the actual discharge angle and the preset target discharge angle is less than the preset deviation threshold, thereby obtaining the optimized modeling result of the belt conveyor. The modules are connected via wired and / or wireless means to enable data transmission between them.

[0065] Through the specific embodiments described above, this invention achieves high-precision modeling and optimization of the hopper transfer point of a belt conveyor by considering the influence of environmental humidity, dynamic friction coefficient allocation, spatiotemporal mesh analysis, stress transmission chain capture, and deviation-driven adjustment. The application of this method not only improves the accuracy and reliability of belt conveyor design but also reduces trial-and-error costs during the design process.

[0066] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0067] It should be noted that all formulas in this manual are calculated by removing dimensions and taking their numerical values. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0068] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. A method for intelligent analysis and optimization of the operating status of a coal mine conveyor, characterized in that, include: Input the geometric structure parameters of the belt conveyor hopper transfer point, the physical property parameters of the granular material, and the environmental humidity parameters into the Bentley software. The physical property parameters include the particle size distribution, density, and initial viscosity of the granular material. Construct a three-dimensional geometric model of the hopper transfer point based on the geometric structure parameters. The physical property parameters of particulate materials are discretized into multiple discrete units to form an initial set of discrete units of particulate materials, and the set of discrete units is mapped to the three-dimensional geometric model. The surface of each discrete unit is divided into multiple micro-grid regions. Under different environmental humidity parameters, the local adhesion properties of each discrete unit are dynamically updated by the changes in the migration path and migration speed of the moisture film between the micro-grid regions. The boundary surface of the hopper transfer point is divided into multiple local boundary regions. The friction coefficient of each local boundary region is dynamically allocated based on the contact frequency and contact pressure between each local boundary region and the discrete unit set of particulate material. Based on the updated local adhesion characteristics and dynamically allocated friction coefficient, the motion process of discrete unit sets of particulate material is simulated in the three-dimensional geometric model. By constructing a spatiotemporal intersection grid, the intersection characteristics of the motion trajectory of discrete units inside the hopper transfer point are captured, and the local flow field behavior of particulate material is extracted. In the extracted local flow field behavior, by constructing stress transmission chains between discrete units, the dynamic fracture and reconstruction process of stress transmission chains is captured, and the accumulation boundary surface of particulate material at the hopper transfer point is determined. The accumulation boundary surface is divided into multiple curvature analysis units. The actual discharge angle of the particulate material is determined by the curvature change rate of the curvature analysis units. Based on the deviation between the actual discharge angle and the preset target discharge angle, a deviation-driven adjustment matrix is ​​constructed. The elements of the adjustment matrix include the adjustment range of the friction coefficient of the local boundary region and the adjustment direction of the local adhesion characteristics of the discrete unit. The three-dimensional geometric model is iteratively updated through the adjustment matrix until the deviation between the actual discharge angle and the preset target discharge angle is less than the preset deviation threshold, thus obtaining the optimized modeling result of the belt conveyor.

2. The intelligent analysis and optimization method for the operating status of a coal mine conveyor according to claim 1, characterized in that, The method of dynamically updating the local adhesion properties of each discrete unit by varying the migration path and velocity of the moisture film between micro-grid regions includes: The surface of each discrete unit is divided into multiple micro-grid regions, and an initial moisture film thickness is assigned to each micro-grid region based on the ambient humidity parameters and the initial viscosity of the particulate material. At different time steps, the migration speed and direction of the water film are determined by the migration path of the water film between adjacent micro-grid regions; The moisture film thickness of each microgrid region is updated based on migration speed and migration direction. Based on the updated moisture film thickness, the local adhesion force of each microgrid region is adjusted through the interaction between the moisture film thickness and the initial viscosity. The local adhesion forces of all tiny grid regions on the surface of each discrete unit are integrated to form the local adhesion properties of the discrete unit.

3. The intelligent analysis and optimization method for the operating status of a coal mine conveyor according to claim 1, characterized in that, The dynamic allocation of the friction coefficient for each local boundary region based on the contact frequency and contact pressure between each local boundary region and the discrete unit set of particulate material includes: The boundary surface of the hopper transfer point is divided into multiple local boundary regions, and the initial friction coefficient is assigned according to the surface roughness of each local boundary region. Within a unit of time, the number of times each local boundary region comes into contact with the discrete unit set of particulate material is counted to form a contact frequency level; Calculate the force distribution in each local boundary region per unit area to form the contact pressure level; Based on the contact frequency level and contact pressure level, the friction coefficient of each local boundary region is adjusted. When the contact frequency level is higher than the preset frequency threshold, the friction coefficient is increased, and when the contact pressure level is higher than the preset pressure threshold, the friction coefficient is decreased. The adjusted friction coefficient is used to form the dynamic friction coefficient for each local boundary region.

4. The intelligent analysis and optimization method for the operating status of a coal mine conveyor according to claim 1, characterized in that, The method of constructing a spatiotemporal intersection grid to capture the intersection characteristics of discrete unit motion trajectories within the hopper transfer point and extracting the local flow field behavior of particulate materials includes: The interior of the hopper transfer point is divided into a spatiotemporal intersection grid, which consists of multiple spatiotemporal units. Each spatiotemporal unit corresponds to a local spatial region and a time segment inside the hopper transfer point. Within each spatiotemporal unit, the motion trajectory of the discrete unit is decomposed into a motion direction vector and a motion velocity vector. The intersection complexity of the spatiotemporal unit is determined by the angle distribution of the motion direction vector and the gradient distribution of the motion velocity vector. In the spatiotemporal intersection grid, spatiotemporal units with intersection complexity exceeding a preset complexity threshold are identified to form high intersection regions; this includes: within each spatiotemporal unit of the spatiotemporal intersection grid, an intersection complexity distribution map is constructed by the concentration of the angle distribution of motion direction vectors and the fluctuation amplitude of the gradient distribution of motion velocity vectors; In the intersection complexity distribution map, the spatiotemporal units corresponding to local peaks are determined by identifying spatiotemporal units with concentrations higher than a preset concentration threshold or fluctuation amplitudes higher than a preset fluctuation amplitude threshold. By integrating the spatiotemporal units corresponding to local peaks, a high convergence region is formed in the spatiotemporal convergence grid; Within the high convergence region, the local flow field behavior of particulate materials is extracted by the spatiotemporal evolution of the motion direction vector and the motion velocity vector.

5. The intelligent analysis and optimization method for the operating status of a coal mine conveyor according to claim 1, characterized in that, The process of constructing stress transmission chains between discrete units, capturing the dynamic fracture and reconstruction process of these chains, and determining the accumulation boundary surface of particulate material at the hopper transfer point includes: In a set of discrete units of particulate material, a stress transmission chain is constructed by identifying the contact points between each discrete unit and its adjacent discrete units. The stress transmission chain consists of a sequence of contact force directions and magnitudes at the contact points. The local stability of the stress transmission chain is determined by the deflection angle of the contact force direction and the fluctuation range of the contact force magnitude at different time steps. In the stress transmission chain, when the deflection angle is greater than a preset angle threshold or the fluctuation amplitude is greater than a preset amplitude threshold, the dynamic break point is identified, and the reconstruction path of the stress transmission chain is determined by the contact point formation between the discrete unit after the dynamic break point and the new adjacent discrete unit. The packing boundary surface of particulate materials is determined by the dynamic fracture and reconstruction path of the stress transmission chain.

6. The intelligent analysis and optimization method for the operating status of a coal mine conveyor according to claim 1, characterized in that, The determination of the actual discharge angle of particulate material by the rate of curvature change of the curvature analysis unit includes: The stacked boundary surface is divided into multiple curvature analysis units, and initial curvature values ​​are assigned according to the local geometry of each curvature analysis unit. At different time steps, the rate of curvature change is determined by the difference in curvature values ​​of the curvature analysis unit between adjacent time steps. In the stacked boundary surface, curvature analysis units with curvature change rates higher than a preset change rate threshold are identified, forming high curvature change regions; Within regions of high curvature variation, the local unloading angle is determined by the angle distribution between the tangent direction and the horizontal plane; The local discharge angles of all high curvature variation regions in the stacking boundary surface are integrated to form the actual discharge angle of the particulate material.

7. The intelligent analysis and optimization method for the operating status of a coal mine conveyor according to claim 1, characterized in that, The constructed deviation-driven adjustment matrix, whose elements include the adjustment magnitude of the friction coefficient in the local boundary region and the adjustment direction of the local adhesion characteristics of the discrete unit, includes: Inside the hopper transfer point, a deviation distribution map is constructed by the deviation between the actual discharge angle and the preset target discharge angle. The deviation distribution map consists of the deviation values ​​of multiple local areas. In the deviation distribution map, the adjustment range of the friction coefficient is determined by the local correlation between the deviation value and the friction coefficient of the local boundary region; In the deviation distribution diagram, the adjustment direction of the local adhesion characteristics is determined by the local correlation between the deviation value and the local adhesion characteristics of the discrete unit. By adjusting the friction coefficient and the direction of local adhesion characteristics, a deviation-driven adjustment matrix is ​​constructed, where the rows of the adjustment matrix correspond to the local boundary region of the hopper transfer point, and the columns correspond to the local adhesion characteristics of the discrete unit set of particulate material. By adjusting the matrix, the update strategy for the three-dimensional geometric model is determined.

8. The intelligent analysis and optimization method for the operating status of a coal mine conveyor according to claim 5, characterized in that, The determination of the stress transmission chain reconstruction path based on the contact point formation between the discrete element and the new adjacent discrete element after the dynamic fracture point includes: Around the dynamic fracture point, the contact point between the discrete element after the dynamic fracture point and the new adjacent discrete element is determined by constructing a local neighborhood. The local neighborhood consists of multiple discrete elements around the dynamic fracture point. Within a local neighborhood, the probability of forming a new contact point is determined by the stability of the contact force direction at the new contact point. The reconstructed path of the stress transmission chain is constructed by using the probability of the formation of new contact points. The reconstructed path consists of a sequence of new contact points and a dynamic adjustment process of the contact force direction.

9. The intelligent analysis and optimization method for the operating status of a coal mine conveyor according to claim 1, characterized in that, The geometric parameters include belt width, belt type, idler type and specifications, drive type, drive / redirection roller specifications, head frame type, tail frame type, support leg type, tensioning device type, hopper inlet size, hopper outlet size, hopper wall thickness, hopper tilt angle, transition point radius of curvature and transition point wall material. The construction of the three-dimensional geometric model also includes: Select the appropriate 3D model of the head frame, tail frame, upper and lower rollers, support legs, and intermediate frame according to the bandwidth. The head and tail frames, as well as the matching rollers and idlers, are automatically matched according to the power and tension. Except for the convex arc section and the material drop point, the upper idler rollers are arranged at 600mm, and the rest are arranged at 1200mm. Except for the convex arc section, the lower idler rollers are arranged at 1200mm, and the rest are arranged at 3000mm. Arrange the outriggers at intervals not exceeding 3000mm; Adjust the height of the head and tail frames and tensioning devices in real time according to their actual heights, and assemble them into their positions; generate a guide chute based on the input drop point position.

10. A coal mine conveyor operation status intelligent analysis and optimization system, used to implement the coal mine conveyor operation status intelligent analysis and optimization method according to any one of claims 1 to 7, characterized in that, include: The parameter acquisition module is used to input the geometric structure parameters of the belt conveyor hopper transfer point, the physical property parameters of the granular material, and the environmental humidity parameters into the Bentley software. The physical property parameters include the particle size distribution, density, and initial viscosity of the granular material, and a three-dimensional geometric model of the hopper transfer point is constructed based on the geometric structure parameters. The discretization module is used to discretize the physical property parameters of particulate materials into multiple discrete units to form an initial set of discrete units of particulate materials, and to map the set of discrete units to the three-dimensional geometric model. The feature update module is used to divide the surface of each discrete unit into multiple micro-grid regions. Under different environmental humidity parameters, the local adhesion characteristics of each discrete unit are dynamically updated by the changes in the migration path and migration speed of the moisture film between the micro-grid regions. The coefficient allocation module is used to divide the boundary surface of the hopper transfer point into multiple local boundary regions, and dynamically allocate the friction coefficient of each local boundary region according to the contact frequency and contact pressure between each local boundary region and the discrete unit set of particulate material. The behavior extraction module is used to simulate the motion process of discrete unit sets of particulate material in the three-dimensional geometric model based on the updated local adhesion characteristics and dynamically allocated friction coefficient. By constructing a spatiotemporal intersection grid, it captures the intersection characteristics of the motion trajectory of discrete units inside the hopper transfer point and extracts the local flow field behavior of particulate material. The boundary determination module is used to determine the accumulation boundary surface of particulate material at the hopper transfer point by constructing stress transmission chains between discrete units in the extracted local flow field behavior, capturing the dynamic fracture and reconstruction process of the stress transmission chains; The angle calculation module is used to divide the accumulation boundary surface into multiple curvature analysis units, and determine the actual discharge angle of the particulate material by the curvature change rate of the curvature analysis units. The iterative optimization module is used to construct a deviation-driven adjustment matrix based on the deviation between the actual discharge angle and the preset target discharge angle. The elements of the adjustment matrix include the adjustment range of the friction coefficient of the local boundary region and the adjustment direction of the local adhesion characteristics of the discrete unit. The three-dimensional geometric model is iteratively updated through the adjustment matrix until the deviation between the actual discharge angle and the preset target discharge angle is less than the preset deviation threshold, thereby obtaining the optimized modeling result of the belt conveyor.

Images