A method and system for generating intelligent leveling schemes for bulk material stockpiles used in the construction of material shed structures.

Abstract

This invention discloses a method and system for generating intelligent leveling schemes for bulk material stockpiles used in the construction of material shed structures, relating to the fields of bulk material stockpiles and steel structure construction technology. The method includes: using a drone to conduct aerial surveying of existing bulk material stockpiles, acquiring three-dimensional spatial data of the stockpiles and generating a digital surface model; constructing a target flat surface model based on the actual surface morphology of existing bulk material stockpiles, combined with the crane's travel route and preset flatness technical indicators for the steel structure stockpiles and assembly areas; automatically generating multiple site treatment schemes based on the three-dimensional spatial data and the target flat surface model using optimization algorithms, taking into account the amount of ore to be processed and the difficulty of the operation; and simultaneously outputting the above schemes and visually displaying them for manual comparison and decision-making. This invention overcomes the shortcomings of traditional manual on-site surveying and planning, such as low efficiency, inaccurate calculations, and difficulty in achieving global optimization, significantly reducing the site leveling cost of adding enclosed material sheds to stockpiles.

Description

Technical Field

[0001] This invention relates to the field of bulk material storage yard construction technology, specifically to a method and system for generating intelligent leveling schemes for bulk material storage yards used in the construction of material shed structures. It is applicable to the intelligent planning and dynamic management of site leveling schemes in the project of adding enclosed material sheds to existing coal mines or ore bulk material storage yards. Background Technology

[0002] In response to environmental protection policies, many open-air bulk material storage yards (such as coal mine and ore storage yards) are gradually undergoing renovation projects by installing enclosed storage sheds. Enclosed storage sheds typically employ large-span steel truss structures. During actual construction, large crawler cranes or truck cranes are required to move and lift materials above the storage yard, and a dedicated truss assembly area must be planned within the site. However, existing bulk material storage yards are in a dynamic stacking and reclaiming state for extended periods, and their surface morphology is often irregularly conical or wavy, with significant elevation changes. This makes it impossible to directly meet the stringent requirements for site flatness needed for the movement of large lifting equipment and the assembly of high-precision steel structures.

[0003] Currently, site leveling planning in the industry mainly relies on manual on-site surveys. Technicians roughly assess elevation differences through visual inspection or conventional surveying instruments, and then formulate excavation, filling, and relocation plans for earthwork (ore) based on experience. This traditional model has the following significant technical drawbacks:

[0004] First, the project is inefficient. Large storage yards cover vast areas with complex terrain, and manual surveying and planning take an extremely long time, severely delaying the progress of the early stages of the project.

[0005] Secondly, construction costs are high. Due to the lack of accurate global three-dimensional volume calculation and allocation path optimization, over-excavation, ineffective filling, or repeated transportation are very likely to occur, resulting in the amount of ore transferred far exceeding the actual demand, which significantly increases the cost of machinery shifts and labor.

[0006] Third, the solution is difficult to achieve global optimization. "Minimizing the amount of ore transferred" and "lowest construction difficulty" (such as avoiding high slopes and reducing human intervention) are usually mutually restrictive multi-objective game processes. Manual planning lacks the support of multi-objective quantitative optimization algorithms, and the final output is often a "feasible solution" rather than an "optimal solution".

[0007] Fourth, the basis for visual decision-making is weak. Traditional two-dimensional drawings or text reports cannot intuitively present the elevation differences and dynamic route adjustments in complex terrain, resulting in managers lacking intuitive and reliable evaluation criteria when faced with multiple options. Summary of the Invention

[0008] This invention provides a method and system for generating intelligent leveling schemes for bulk material stockpiles used in the construction of material shed structures, in order to solve the problems of low efficiency of manual planning, rough earthwork volume calculation, and inability to achieve global optimization in the prior art.

[0009] To achieve the above objectives, the present invention adopts the following technical solution: a method for generating a smart leveling scheme for bulk material storage yards used in the construction of material shed structures, executed by a computing device, comprising the following steps:

[0010] S1. UAV flight mapping steps:

[0011] A systematic aerial survey of existing bulk material stockpiles is conducted using unmanned aerial vehicles (UAVs) equipped with high-precision mapping sensors. This process acquires three-dimensional spatial data of the stockpile area, generating a digital surface model (DSM) or three-dimensional point cloud data. The mapping sensors can be LiDAR (Light Detection and Ranging) or high-precision visible light cameras. For flight parameters, the forward overlap rate is no less than 70%, and the lateral overlap rate is no less than 60%, ensuring complete coverage of the point cloud data. The spatial resolution of the obtained data is no less than 5 cm, and the elevation accuracy is better than 10 cm, meeting the requirements for subsequent refined earthwork volume calculations and slope analysis.

[0012] S2. Steps for constructing the target flat surface:

[0013] Based on the engineering design drawings of the enclosed material shed, the path of the crane's travel track, the steel component assembly area and the boundary of the stockpile are extracted, and a target flat surface model is constructed in combination with the actual shape of the existing bulk material stockpile.

[0014] The target flat surface refers to the geometric surface that the crane travel area and steel structure construction area should achieve after leveling. Specific construction methods include:

[0015] (1) Target flat surface of the crane travel area: The target flat surface is set as a plane or slope with a gradient less than or equal to the preset gradient threshold α, and the elevation of the plane or slope meets the safety distance requirements for crane travel (i.e., the crane will not overturn or cause structural interference due to excessive ground undulations during travel). The value of α ranges from 0° to 5°, and the specific value is determined according to the climbing ability and safety parameters of the crane model. For crawler cranes that are sensitive to slope, the value of α is relatively small (e.g., 0° to 2°); for tire cranes with strong climbing ability, the value of α can be appropriately relaxed.

[0016] (2) Target flat surface of the truss construction (assembly) area: The target flat surface is set as a horizontal surface or a designed inclined surface with an elevation error within the preset error range β. The typical value of β is ±50mm to ±150mm, and the specific value is determined according to the accuracy requirements of steel structure assembly. For the main chord joint of a large-span truss, the value of β is stricter (e.g., ±50mm); for the assembly of secondary structures, the value of β can be appropriately relaxed to ±150mm.

[0017] Using the 3D point cloud data obtained in step S1 as a spatial reference, a theoretical target flat surface model is generated based on the aforementioned indicators. The system automatically performs Boolean operations, that is, overlays and analyzes the current digital surface model and the target flat surface model in 3D space to extract the volume difference between the current surface and the target flat surface. The area where the current surface is higher than the target flat surface is the excavation area, and the area where the current surface is lower than the target flat surface is the fill area. The sum of the volumes of the two is the theoretical amount of ore that needs to be excavated, filled, and transferred.

[0018] S3. Intelligent Solution Generation Steps:

[0019] Based on the existing 3D data and the target flat surface model, a built-in multi-objective optimization algorithm is invoked for global optimization calculation. This multi-objective optimization algorithm can employ heuristic optimization algorithms such as genetic algorithms or particle swarm optimization.

[0020] To eliminate the interference of different physical dimensions and their magnitudes on the multi-objective optimization results, the optimization algorithm is solved using a normalized bi-objective function:

[0021]

[0022] in, This represents the volume of ore to be transferred according to the corresponding scheme.

[0023] The initial estimated total volume of the ore to be processed, calculated in step S2;

[0024] This is an indicator of the geometric complexity of the path.

[0025] This is the preset maximum allowable complexity index;

[0026] and The preset weighting coefficients, and + = 1;

[0027] The path geometric complexity index is calculated based on at least the following factors:

[0028] (a) Local maximum slope on the crane's travel path: This value reflects the maximum slope risk faced by the crane during its travel. The greater the slope, the worse the crane's travel stability, requiring more precise site preparation and increasing the operational difficulty. The system extracts the elevation values ​​of each grid cell traversed by the crane's travel trajectory in the candidate schemes, calculates the slope between adjacent cells, and takes the maximum value as the local maximum slope.

[0029] (b) Percentage of area requiring manual leveling: Mechanical leveling (such as bulldozers and graders) is suitable for large-area, high-efficiency operations, but in areas with abrupt terrain changes or near crane outrigger support points, machinery may be difficult to reach or cannot guarantee accuracy, requiring manual assistance for fine leveling. The higher this percentage, the more limited mechanized operations are, and the higher the labor and time costs.

[0030] (c) The degree of interference between leveling operations and normal production operations in the stockyard: For stockyards where production and construction are carried out simultaneously, leveling operations may occupy production channels or affect stacking and reclaiming operations. The longer the interference lasts, the greater the impact on the normal operation of the stockyard. In this invention, the degree of interference is quantified by the duration of the interference.

[0031] In terms of the optimization mechanism, by adjusting the weight coefficients in the multi-objective optimization algorithm ( and This can generate a set of optimal solutions with different tendencies. The set of solutions includes at least:

[0032] Appropriate detour plan, prioritizing material saving: the goal is to minimize the amount of ore transferred (i.e. (The value is relatively large). This scheme allows the crane path to detour around high-slope areas to obtain a smaller amount of ore excavation and filling.

[0033] Incremental transfer plan, prioritizing labor saving: Prioritizing the lowest operational difficulty (i.e.) (The value is relatively large). This solution treats the site to a more strictly level state. Although it increases the amount of ore to be transferred, it effectively reduces the risk of crane travel on slope and the amount of manual leveling work.

[0034] The optimal solution is to balance the amount of ore transferred with the difficulty of the operation (i.e., and All values ​​are taken as moderate values ​​to minimize the overall cost (ore transfer cost and operational difficulty cost).

[0035] Each plan includes detailed information on: the amount of ore to be transferred or leveled in each processing area, the assessment results of the operational difficulty, and the corresponding estimated construction period and cost data.

[0036] Furthermore, for complex operating conditions involving both production and construction, the multi-objective optimization algorithm can incorporate the requirement of "reserving at least one through temporary access road for yard production and transportation vehicles" as a rigid geometric boundary constraint into the optimization model during optimization calculations. The candidate solution set includes different temporary access road layout schemes, and the system recommends the access road scheme with the least overall impact based on the incremental transfer volume and production interference duration corresponding to each scheme.

[0037] S4. Solution Output and Decision-Making Steps:

[0038] The various candidate solutions generated in step S3 are visualized from multiple dimensions. Specific visualization methods include:

[0039] (1) Display the height difference distribution between the existing bulk material loading status and the target flat surface using a color contour map or three-dimensional perspective view; wherein, the contour map uses color gradation to reflect the cut and fill thickness, with warm red indicating the high areas that need to be excavated and cool blue indicating the low-lying areas that need to be backfilled.

[0040] (2) Mark the direction and amount of ore transfer for each scheme with arrows or color blocks; the thickness of the arrows or the size of the color blocks are proportional to the amount of transfer, so that managers can intuitively judge the direction and scale of material flow.

[0041] (3) Output a data comparison table containing the transfer volume, maximum slope of the path, operation difficulty score, expected construction period, and expected cost of the corresponding scheme; the comparison table compares the schemes side by side according to a unified evaluation dimension (transfer volume, difficulty score, maximum slope, construction period, cost, etc.).

[0042] Based on the aforementioned quantitative data, the system automatically calculates and recommends the optimal comprehensive solution. Users can adjust the weighting coefficients through a visual interactive interface (e.g., increasing them when the project schedule is extremely tight). The weights are adjusted so that the system tends to output solutions with lower difficulty levels. The system responds to the weight adjustment command by recalculating in real time and updating the solution display.

[0043] After final confirmation by the human, the system responds to the human selection confirmation command and exports a construction guidance document with one click, which includes a site layout plan (including cut and fill boundary lines), longitudinal and cross-sectional views (arranged according to the design spacing, showing the cut and fill balance relationship), and a bill of quantities (specifying the cut volume, fill volume, and transportation distance by area), which serves as the final site leveling construction plan.

[0044] S5. Implementation and monitoring steps of the solution (as a preferred implementation method):

[0045] Using drones, on-site re-measurements are conducted at preset intervals to obtain updated three-dimensional spatial data of the stockpile during the construction phase, and to generate a measured digital surface model containing real-time terrain features.

[0046] Calculate the real-time volume difference between the measured digital surface model and the target flat surface model; this difference is the amount of ore remaining to be processed.

[0047] The real-time volume difference is returned to step S3 as an input parameter, and the multi-objective optimization algorithm is called again. Based on the updated current data, the optimal processing strategy for the remaining working area is recalculated, and the leveling control strategy for the remaining working area is updated to form a closed-loop digital management and control system.

[0048] To achieve the above method, the present invention also provides a smart leveling scheme generation system for bulk material storage yards used in the construction of material shed structures, comprising:

[0049] Data acquisition module: used to receive three-dimensional spatial data of the yard area obtained by UAV flight mapping; this module has a multi-source data interface and is compatible with point cloud data generated by lidar point cloud data and visible light imagery.

[0050] Model building module: connected to the data acquisition module, used to generate a digital surface model based on the three-dimensional spatial data, and to build a target flat surface model based on engineering design parameters and current status data of the stockpile; this module has a built-in Boolean operation engine, which can automatically calculate the volume difference between the current status and the target surface.

[0051] Solution generation module: Connected to the model building module, it is used to calculate multiple candidate site treatment solutions with the objectives of minimizing ore transfer and minimizing operational difficulty through optimization algorithms; this module has a built-in genetic algorithm / particle swarm optimization solver and supports flexible configuration of weight coefficients.

[0052] Decision output module: Connected to the solution generation module, it is used to visualize the multiple candidate solutions and respond to manual selection instructions in a multi-objective decision-making manner to output the optimal solution; this module has a 3D scene rendering engine and automatic data report generation function.

[0053] The modules are connected via a data bus to enable data exchange and collaborative computation.

[0054] Compared with the prior art, the present invention has the following beneficial effects:

[0055] 1. Achieving global optimization of ore allocation and effectively reducing site leveling costs. This invention introduces a normalized bi-objective optimization function, effectively eliminating dimensional differences and algorithmic biases between volume and complexity variables. While strictly meeting the technical requirements for large crane movement and truss assembly, it effectively reduces ineffective ore transfer, handling, and manual intervention, resulting in actual savings of 15%–30% in site leveling costs.

[0056] 2. Improve the efficiency of site survey and scheme preparation. High-density 3D aerial mapping using UAVs replaces traditional manual on-site surveys. Combined with an intelligent computing engine, Boolean operations and algorithm optimization are performed directly based on the Digital Surface Model (DSM). This eliminates the time-consuming and labor-intensive limitations of traditional manual experience-based assessment and extensive calculations, shortening the project preparation and scheme planning cycle.

[0057] 3. Provides multi-dimensional visualization support to achieve scientific and precise engineering decision-making. This system outputs optimal solutions covering different preferences. Through 3D perspective views, contour line color-coded maps, and multi-dimensional indicator comparison tables, it transforms abstract spatial data into intuitive visual references, effectively assisting management in quickly identifying the optimal implementation plan.

[0058] 4. It possesses strong flexibility and adaptability to various engineering scenarios. This invention allows users to adjust the weight coefficients in the objective optimization function based on dynamic constraints at the engineering site (such as sudden shortening of the construction period or budget tightening). The system can iterate to find new optimal solutions, fitting the complex and ever-changing actual construction environment. Furthermore, by introducing rigid geometric constraints such as production support channels, this invention can effectively cope with complex conditions involving simultaneous production and construction.

[0059] 5. Establish a closed-loop digital model to achieve digital control and record-keeping throughout the entire construction process. From the initial site survey and scheme generation to the dynamic re-measurement and verification by drones and strategy iteration during construction, the excavation and filling work volume and progress ledger are digitally recorded throughout the entire process, providing accurate and complete data support for later settlement and auditing. Attached Figure Description

[0060] Figure 1 This is a flowchart illustrating the overall process of the present invention.

[0061] Figure 2 This is a schematic diagram of the UAV flight path, showing the outline of the yard area and the UAV's serpentine scanning flight path, which covers the entire yard area.

[0062] Figure 3 This is a schematic diagram of a 3D point cloud of a stockpile obtained by UAV aerial surveying. The point cloud data presents the surface morphology of the stockpile in 3D coordinate form, and the elevation information of each point is represented by color mapping. Warm colors represent areas with higher elevation, and cool colors represent areas with lower elevation.

[0063] Figure 4 This is a top-down schematic diagram of the yard and crane travel area, showing the layout of the yard area, the crane travel path, the relative position of the truss assembly area, and the positions of the AA and BB section lines.

[0064] Figure 5 For along Figure 4 The cross-sectional diagram of the AA profile line shows that the solid line represents the current surface curve of the stockpile, the dashed line represents the target flat surface, and the diagonal filled area represents the excavation area and the filling area.

[0065] Figure 6 For along Figure 4 The cross-sectional diagram of the BB profile shows that the solid line represents the current surface curve of the stockpile, the dashed line represents the target flat surface, and the diagonal filled area represents the excavation area and the filling area.

[0066] Figure 7 This is a 3D visualization comparison diagram of various site treatment solutions, in which... Figure 7 (a) is the elevation difference color-coded diagram for Option 1, "Material Saving Priority". Figure 7 (b) is the elevation difference color-coded map for Option 2, "Prioritizing Labor Saving". Figure 7 (c) is the elevation difference color-filled map of Scheme 3 "Comprehensive Optimization". The ore transfer amount of the corresponding scheme is marked below each map.

[0067] Figure 8 This is a schematic diagram of the solution decision output interface. The solution comparison table is displayed in the form of a simulation software interface. The columns of the table are, in order, solution number, ore transfer volume, operation difficulty score, maximum slope of crane path, estimated construction period, estimated cost and recommendation level. There are operation buttons at the bottom of the interface. Detailed Implementation

[0068] The following is in conjunction with the appendix Figures 1 to 8 The present invention provides a more detailed description of the specific implementation method and system for generating intelligent leveling schemes for bulk material storage yards used in the construction of material shed structures.

[0069] Example 1: Installation of Enclosed Material Sheds at a Coal Mine Stockyard

[0070] Project Background: A coal mining company plans to install a large-span steel-structured enclosed material shed at its existing open-pit coal yard. The shed will have a span of 120m and a length of 300m. Construction requires a QY80-ton truck crane to operate above the coal pile, and a truss assembly area (30m x 40m) needs to be set up on one side of the shed. The current coal pile is an irregular cone-shaped accumulation, with a maximum height of approximately 15m and a total area of ​​approximately 35,000m². 2 The project requires site leveling to be completed and steel structure installation to begin within 6 months.

[0071] Step S1: UAV flight mapping

[0072] like Figure 2 As shown, a DJI Matrice 300 RTK drone, equipped with a Zenmuse L1 LiDAR sensor, was used to perform a serpentine scanning flight mapping of the coal pile area. The flight altitude was 80m, the flight speed was 8m / s, the forward overlap was 75%, the lateral overlap was 65%, and the LiDAR scanning frequency was 160kHz. The flight path covered the entire stockpile area, ensuring comprehensive data collection. Approximately 120 million point cloud data points were acquired.

[0073] The raw point cloud data were preprocessed as follows: (a) statistical filtering was used to remove outliers; (b) ground points and non-ground points were separated based on the cloth simulation filtering (CSF) algorithm; (c) a DSM digital surface model with a resolution of 5 cm was generated using the thin plate spline interpolation (TPS) algorithm.

[0074] like Figure 3 As shown, the generated 3D point cloud scatter plot presents the surface morphology of the stockpile in 3D coordinate form, and uses color mapping to represent the elevation information of each point—warm colors (red-orange) represent areas with higher elevations, and cool colors (blue-green) represent areas with lower elevations. Accuracy verification shows that the elevation error is better than 8cm, meeting the requirements for subsequent refined earthwork volume calculations and slope analysis.

[0075] Step S2: Constructing the target flat surface

[0076] The following parameters are extracted from the engineering design drawings: the crane travel path is within 15m on both sides of the longitudinal centerline of the material shed, the safety slope threshold is set to 0°~2° (preferably α=2°), and the road width is 6m; the truss assembly area is located on the west side of the material shed, the design elevation benchmark is +2.5m (relative to the lowest point of the stockpile), and the assembly error threshold is set to β=±100mm.

[0077] like Figure 4 As shown, the yard area is an irregular polygon. The figure marks the planar positional relationship of the yard boundary, the crane travel path (extending along the longitudinal centerline), and the truss assembly area (located in the rectangular area on the west side of the yard). The positions of the two section lines AA and BB are also shown.

[0078] The system uses the DSM current status data as a spatial reference and constructs the target flat surface model through the following sub-steps:

[0079] Sub-step S2-1: Generate the target surface of the crane's travel area—Along the centerline of the crane's travel path, generate a horizontal strip plane with a slope of 0° with a bandwidth of 6m (considering that the climbing ability of the QY80-ton truck crane is 3°, we take α=2° to leave a safety margin).

[0080] Sub-step S2-2: Generate the target surface of the truss assembly area—Generate a theoretical horizontal surface with an elevation of +2.5m and an error of ±100mm within a 30m×40m rectangular area.

[0081] Sub-step S2-3: Perform Boolean operations—align the current DSM with the target flat surface model in three-dimensional space, calculate the elevation difference grid by grid, and obtain the cut-fill volume distribution map.

[0082] like Figure 5As shown, along cross section AA (transverse profile), the solid line represents the current surface curve of the stockpile obtained based on the 3D point cloud model, and the red dashed line represents the designed target flat surface. Between the solid and dashed lines, areas where the current surface is higher than the target flat surface are represented by diagonal fill lines (excavation areas), and areas where the current surface is lower than the target flat surface are represented by diagonal fill lines (fill areas). It can be clearly seen from the figure that there is a large area of ​​high stockpiles in the center of the stockpile that needs to be excavated, and low-lying areas on both sides that need to be backfilled.

[0083] like Figure 6 As shown, the cross-section BB (longitudinal profile) further illustrates the excavation and filling distribution along the length of the stockpile. A comprehensive analysis of the two sections indicates that the stockpile exhibits an overall conical shape, with a high center and low perimeter. Excavation is primarily concentrated at the top of the stockpile, while filling is mainly distributed in the peripheral areas.

[0084] Preliminary calculations indicate that the initial estimated total volume V0 of the ore to be processed, meeting the most stringent flatness requirements, is approximately 5000 m³. 3 .

[0085] Step S3: Intelligent Solution Generation

[0086] The system invokes a multi-objective optimization algorithm, sets a maximum allowable complexity index, and extracts local maximum slope and the proportion of manually leveled area as path geometric complexity indicators. The weight coefficients in the objective function are dynamically adjusted. (Volume weight) and (Complexity weights) The system optimizes within a floating range that satisfies the slope and elevation safety thresholds, and outputs solutions with the following three typical biases:

[0087]

[0088] Step S4: Solution Output and Decision (see...) Figure 7 , Figure 8 )

[0089] The system presents three solutions using a 3D visualization interface. For example... Figure 7 As shown, Figure 7 (a) is the elevation difference color-coded diagram for Option 1, "Material Saving Priority". Figure 7 (b) is the elevation difference color-coded map for Option 2, "Prioritizing Labor Saving". Figure 7(c) This is the elevation difference color-coded map for Scheme 3, the "optimal overall solution." All three sub-maps use red-blue contour lines to show the cut-and-fill distribution for each scheme. Red areas (warm colors) represent the elevated portions (cut areas) that need to be removed, while blue areas (cool colors) represent the low-lying portions (fill areas) that need to be backfilled. Comparing the three sub-maps reveals that: Scheme 1 has a more concentrated cut-and-fill distribution, retaining some high-slope areas without treatment, thus minimizing the transfer volume but resulting in steeper local slopes; Scheme 2 has the widest leveling area and the most uniform cut-and-fill distribution, resulting in the lowest operational difficulty but a significantly increased transfer volume; Scheme 3's cut-and-fill distribution falls between the two, controlling the transfer volume while reducing operational difficulty. The ore transfer volume for each scheme is labeled below each map for easy comparison.

[0090] The right sidebar is Figure 8 The data comparison table shown presents the core indicators of each scheme side by side. Figure 8 The interface displays the decision-making output for the proposed solutions. The top of the interface features a title bar, while the main central area displays a table comparing the proposed solutions. The table has seven columns: solution number, ore transfer volume (m³), etc. 3 The system displays the task difficulty score, maximum slope of the crane path, estimated construction period (days), estimated cost (ten thousand yuan), and recommendation level, with the recommendation level automatically assigned by the system based on the comprehensive score. The bottom of the interface features three operation buttons: "Recalculate," "Export Results," and "Confirm Output." The system highlights the recommended solution (Solution 3).

[0091] After viewing the detailed information of each solution through the interactive interface, the user believes that Solution 3 (the best overall solution) achieves the best balance between cost (198,000 yuan) and construction period (10 days), and the difficulty score (58 points) is within an acceptable range. Finally, the user clicks to confirm and select Solution 3 as the implementation plan.

[0092] In response to a manual selection confirmation command, the system can export the following construction guidance documents with a single click:

[0093] (a) 1:500 site leveling plan (CAD format.dwg), with cut and fill boundary lines and zone numbers marked;

[0094] (b) Typical cross-sectional views (a total of 5 cross-sections along AA, BB and three other auxiliary profile lines) show the cut-fill balance relationship;

[0095] (c) Detailed bill of quantities (Excel format), listing excavation volume, fill volume, and average transport distance by region;

[0096] (d) Construction schedule (Gantt chart format), with each process arranged by day.

[0097] Step S5: Construction Monitoring (Optional)

[0098] On the fifth day of construction, the first resurvey was conducted using a drone. The flight parameters for the resurvey were consistent with the initial survey. A grid-by-grid differential comparison was performed between the measured DSM and the design DSM, calculating that approximately 3100 m³ of excavation and filling work had been completed. 3 The remaining workload is approximately 1900m. 3 The system automatically generates a progress log, showing the actual progress at 62.5%, slightly ahead of the planned progress (50%).

[0099] The system will use the remaining volume difference (1900m) 3 The algorithm returns to step S3 as input parameters, and calls the optimization algorithm to recalculate the leveling control strategy for the remaining working area. Since the remaining area is mainly distributed in the fill area with low elevation, the algorithm suggests adjusting the focus of subsequent work to "transporting ore from the nearby excavation area for backfilling to reduce long-distance transportation", and updates the progress guidance log for the remaining working area.

[0100] During construction, the system underwent drone retesting every 5 days, with a total of 5 iterations and updates. The final actual construction period was 11 days, and the actual cost was 201,000 yuan, which is basically consistent with the budget of Option 3 (10 days / 198,000 yuan). After leveling, the site was tested and found that the maximum slope of the crane travel area was 0.9°, and the elevation error of the truss assembly area was within ±80mm, which fully met the construction technical requirements.

[0101] Example 2: Renovation Project of Enclosed Shed for Iron Ore Stockpile

[0102] Project Background: A steel company is installing enclosed storage sheds in its existing open-pit iron ore storage yard, which covers an area of ​​approximately 50,000 m². 2 The ore piles reached a maximum height of 20 meters. Construction was carried out using crawler cranes, with the assembly area located on the east side of the stockpile. In this case, the ore pile shape was more complex, with multiple independent ore piles and work access routes.

[0103] The difference from Example 1 is that when executing step S3 (intelligent scheme generation), the system uses "the interference duration between leveling operations and normal production operations in the stockpile" as the core weighting term of the path geometric complexity index. To reduce this index, the optimization algorithm automatically incorporates "reserving at least one through temporary operation channel for ore transport vehicles" as a hard geometric boundary constraint into the model during the solution process.

[0104] Ultimately, the system not only output the optimal solutions for the assembly and leveling areas, but also automatically planned two candidate production-preserving channels that would not affect the safety of the cranes. By comparing the incremental transfer volume and interference duration corresponding to the two channels, the system recommended the channel solution with the least overall impact, effectively ensuring uninterrupted production at the iron ore stockpile during the renovation and construction period. This fully demonstrates the high adaptability of this invention in dealing with complex dynamic constraints.

[0105] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principle of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A method for generating a smart leveling scheme for bulk material storage yards used in the construction of material shed structures, characterized in that, Performed by a computing device, including the following steps: S1. UAV aerial surveying steps: Use a UAV equipped with surveying sensors to conduct aerial surveying of the existing bulk material stockpile, obtain three-dimensional spatial data of the stockpile area, and generate digital surface models or three-dimensional point cloud data. S2. Target flat surface construction steps: Based on the engineering design parameters of the addition of enclosed material sheds to the stockpile, determine the planar position and site flatness requirements of the crane travel area, steel structure stockpile and assembly construction area, and construct the target flat surface model based on the actual shape data of the existing bulk material stockpile in the stockpile. The target flat surface is the geometric surface that the crane travel area and steel structure construction area should achieve after flattening. S3. Intelligent generation steps of the scheme: Based on the three-dimensional spatial data and the target flat surface model, the path geometric complexity index is extracted, and combined with the amount of ore to be transferred, the candidate scheme set for site treatment is calculated through a multi-objective optimization algorithm; wherein, each scheme includes the amount of ore to be transferred or flattened in each treatment area and the corresponding operation difficulty assessment parameters. S4. Solution Output and Decision-Making Steps: Visualize the generated candidate solution set and use the minimum amount of ore transferred and the lowest overall operational difficulty as the basis for multi-objective decision-making. In response to the manual selection and confirmation instruction, output the final site leveling construction solution.

2. The method for generating a smart leveling scheme for bulk material storage yards for material shed structure construction according to claim 1, characterized in that, In step S2, the method for constructing the target flat surface model includes: The target flat surface of the crane's travel area is defined as a plane or slope with a slope less than or equal to a preset slope threshold α, and the elevation of the plane or slope meets the safe travel distance requirements of the crane; wherein, the value of α ranges from 0° to 5°. The target flat surface of the steel structure storage yard and assembly construction area is set as a horizontal surface or a designed inclined surface with an elevation error within a preset error range β; wherein, the value of β ranges from ±50mm to ±150mm. The volume difference between the target flat surface corresponding to the crane travel area and the steel structure stockpile and assembly construction area and the actual shape of the existing bulk material stockpile is obtained as the amount of ore that needs to be transferred.

3. The method for generating a smart leveling scheme for bulk material storage yards for material shed structure construction according to claim 1, characterized in that, In step S3, the multi-objective optimization algorithm is a normalized bi-objective function: in: This represents the volume of ore to be transferred according to the corresponding scheme. The initial estimated total volume of the ore to be processed, calculated in step S2; This is a measure of the geometric complexity of the path. This is the preset maximum allowable complexity index; and The preset weighting coefficients, and + = 1; The path geometric complexity index is calculated based on at least the following factors: the maximum local slope on the crane's travel path, the proportion of the area requiring manual leveling, and the degree of interference between leveling operations and normal production operations in the storage yard.

4. The method for generating a smart leveling scheme for bulk material storage yards for material shed structure construction according to claim 1, characterized in that, In step S3, by adjusting the weight coefficients in the multi-objective optimization algorithm, a set of optimal solutions with different tendencies is generated. The set of solutions includes at least: a moderate detour solution with the minimum ore transfer volume as the priority objective, an incremental transfer solution with the minimum operational difficulty as the priority objective, and a comprehensive optimal solution that balances the ore transfer volume and operational difficulty. Each solution is associated with corresponding estimated construction period and cost estimation data.

5. The method for generating a smart leveling scheme for bulk material storage yards for material shed structure construction according to claim 1, characterized in that, In step S4, the visualization methods of the solution include: Use color contour maps or 3D perspective views to show the height difference distribution between the existing bulk material loading status and the target flat surface; Use arrows or color blocks to indicate the direction and amount of ore transfer for each scheme; The output includes a data comparison table of the corresponding scheme's transfer volume, maximum path gradient, operation difficulty score, estimated construction period, and estimated cost.

6. The method for generating a smart leveling scheme for bulk material storage yards for material shed structure construction according to claim 1, characterized in that, It also includes the implementation monitoring steps of the solution: Acquire updated 3D spatial data of the stockpile during the construction phase and generate a measured digital surface model containing real-time terrain features; Calculate the real-time volume difference between the measured digital surface model and the target flat surface model; The real-time volume difference is used as an input parameter to return to step S3, and the multi-objective optimization algorithm is called again to update the leveling control strategy of the remaining working area.

7. The method for generating a smart leveling scheme for bulk material storage yards for material shed structure construction according to claim 1, characterized in that, In step S1, the forward overlap rate of the UAV flight mapping is not less than 70%, the lateral overlap rate is not less than 60%, and the elevation accuracy of the three-dimensional spatial data is better than 10cm.

8. The method for generating a smart leveling scheme for bulk material storage yards for material shed structure construction according to claim 1, characterized in that, In step S3, the multi-objective optimization algorithm is a genetic algorithm or a particle swarm optimization algorithm.

9. The method for generating a smart leveling scheme for bulk material storage yards for material shed structure construction according to claim 1, characterized in that, In step S3, when the multi-objective optimization algorithm performs optimization calculations, it incorporates "reserving at least one through temporary operation channel for yard production and transportation vehicles to pass" as a hard geometric boundary constraint into the optimization model; the candidate scheme set includes different temporary channel layout schemes, and the system recommends the channel scheme with the least comprehensive impact based on the transfer volume increment and production interference duration corresponding to each scheme.

10. A smart leveling scheme generation system for bulk material storage yards used in the construction of material shed structures, characterized in that, include: The data acquisition module is used to receive three-dimensional spatial data of the yard area obtained by UAV flight mapping; The model building module, connected to the data acquisition module, is used to generate a digital surface model based on the three-dimensional spatial data, and to build a target flat surface model based on engineering design parameters and current stockpile data. The scheme generation module, connected to the model building module, is used to calculate multiple candidate site treatment schemes through optimization algorithms with multiple objectives, including minimizing ore transfer and minimizing operational difficulty. The decision output module, connected to the scheme generation module, is used to visualize multiple candidate schemes and output the optimal scheme in response to manual selection instructions in a multi-objective decision-making manner. The modules are connected via a data bus to enable data exchange and collaborative computation.

Images