Automatic distribution method and system of sand and gravel yard unloading trolley

Abstract

This invention relates to the field of material distribution control technology, providing an automatic material distribution method and system for a sand and gravel stockpile unloading trolley. It acquires a point cloud of the stockpile after filtering out fixed structural points and completing the data, achieving global and accurate perception of the stockpile's three-dimensional morphology. Using the constraint that the surface height of each grid cell does not exceed a preset safety height limit as a hard constraint, any candidate sites that might lead to local exceedances are eliminated before unloading. The global termination condition is the absence of candidate sites that meet the safety constraints. This transforms passive, reactive manual detection into proactive, pre-emptive control, eliminating safety hazards such as chute blockage and motor overload at the source. Virtual unloading simulation is performed for each candidate unloading site, using the material's angle of repose as a physical constraint to predict the stockpile morphology after unloading. The standard deviation of the expected height distribution is calculated as a quantitative index of uniformity, achieving quantitative optimization decision-making with the goal of optimal global uniformity, significantly improving stockpile uniformity while ensuring safety.

Description

Technical Field

[0001] This invention relates to the field of material control technology, and in particular to an automatic material placement method and system for a sand and gravel stockpile unloading trolley. Background Technology

[0002] The material placement operation at a sand and gravel stockpile refers to the production process of placing materials into the silo using unloading trolleys. Currently, most sand and gravel mine stockpiles in China use manually operated unloading trolleys for material placement. On-site operators visually observe the height of the material pile below the unloading trolley and manually adjust the movement of the unloading trolley using buttons on the control box based on changes in the pile height.

[0003] The manual handling of fabric has the following problems: (1) Operators cannot accurately grasp the surface morphology distribution of materials in the entire silo by visually judging the height of the material pile at the unloading point. If the material pile is too high and not detected in time, it will cause problems such as material blockage in the unloading chute of the unloading trolley, increased walking pressure, and motor overload, which will affect production and pose safety hazards.

[0004] (2) Manual fabric laying relies entirely on the operator's personal experience, sense of responsibility and judgment. There are skill differences between different operators, and the same operator may experience fluctuations in fabric quality due to fatigue and other factors at different times, making it difficult to ensure the uniformity of the material pile.

[0005] (3) The working environment is harsh. The sand and gravel yard is a high dust and high noise working environment. Operators are exposed to it for a long time, which is not conducive to occupational health.

[0006] In view of this, it is necessary to propose an automatic material placement method and system for sand and gravel stockpile unloading trolleys to solve or at least alleviate the above-mentioned defects. Summary of the Invention

[0007] The main objective of this invention is to provide an automatic material placement method and system for a sand and gravel stockpile unloading trolley, in order to solve the technical problems of untimely detection of material pile height and insufficient uniformity of material pile in the prior art when using manual material placement.

[0008] To achieve the above objectives, the present invention provides an automatic material placement method for a sand and gravel stockpile unloading trolley, comprising the following steps: S1, acquire the ground reference height of the silo and the point cloud of the current material pile; wherein, the point cloud of the material pile is three-dimensional point cloud data that has been filtered out of fixed structural points of the silo and has been completed; S2, determine at least one candidate unloading site; S3, Perform a virtual unloading simulation for each candidate unloading site to generate the expected material pile point cloud corresponding to each candidate unloading site, and calculate the expected height distribution standard deviation of the expected material pile point cloud; S4. Among the candidate unloading sites that meet the preset constraints, select the one with the smallest standard deviation of the expected height distribution as the target unloading site. The preset constraints are: after the expected material pile point cloud is divided into grids, the surface height of the material pile in each grid cell does not exceed the preset safe height limit at that grid cell. S5, control the unloading trolley to move to the target unloading point for unloading. When the height of the material pile at the target unloading point reaches the preset single-point unloading height threshold or the unloading amount reaches the preset single unloading amount, stop the unloading at the target unloading point and return to step S1. Until there are no candidate unloading points that meet the preset constraints in step S4, terminate the entire material laying process.

[0009] Preferably, step S1 includes the following steps: S11, acquire the template point cloud collected in the empty warehouse state, and filter the ground point cloud from the template point cloud according to the height threshold, and use it as the ground reference height of the warehouse. S12, acquire the target point cloud collected under the current material pile state, perform spatial transformation on the template point cloud to align the template point cloud with the target point cloud, and obtain the registered template point cloud; S13, for each point in the target point cloud, find the nearest neighbor of the point in the registered template point cloud, calculate the Euclidean distance between the point and the nearest neighbor, retain the points with Euclidean distance greater than the preset distance threshold as material pile points, and filter out the remaining points as fixed structure points to obtain the preliminary material pile point cloud. S14, Identify the void regions in the preliminary material pile point cloud, and fill and complete the void regions using a constraint interpolation method based on the physical properties of the material to obtain the material pile point cloud.

[0010] Preferably, step S2 includes the following steps: S21, extract the cross-sectional height curve from the point cloud of the material pile along the moving direction of the unloading trolley; wherein, the cross-sectional height curve reflects the highest height of the material pile surface at each sampling point along the moving direction, and the sampling point is a point divided along the moving direction of the unloading trolley by a preset sampling interval; S22, the sampling points on the cross-sectional height curve whose height is lower than the average height of the entire cross-sectional height curve and whose height is lower than the immediately preceding and following sampling points are determined as the candidate unloading points.

[0011] Preferably, the step S21 of extracting the cross-sectional height curve along the moving direction of the unloading trolley includes the following steps: S211, along the moving direction of the unloading trolley, determine each sampling position point according to the preset sampling interval; S212, For each sampling location point, take the maximum value of the height of each point in the cross section perpendicular to the moving direction of the material pile point cloud at that sampling location point, and take it as the cross section height corresponding to that sampling location point; S213, the cross-sectional heights corresponding to each sampling location point are connected sequentially according to the moving direction to form the cross-sectional height curve.

[0012] Preferably, step S3 includes the following steps: S31, divide the material pile point cloud into horizontal plane grids to obtain several grid units. Take the maximum height of each material pile point in each grid unit as the material pile surface height of the grid unit. Subtract the material silo ground reference height at the grid unit from the material pile surface height to obtain the net height of the material pile of the grid unit. S32, with the candidate unloading site as the center, construct an accumulation body according to the preset single unloading amount and material repose angle, and determine the accumulation height of the accumulation body at each grid cell; S33, for each grid cell, take the larger value between the stacking height and the surface height of the material pile as the height of the superimposed grid cell, and form a superimposed height matrix; S34, perform slope constraint correction on the superimposed height matrix, reduce the part of the local slope that exceeds the material repose angle to meet the limit value of the material repose angle, and obtain the corrected height matrix; S35, convert the horizontal position coordinates and corresponding corrected heights of each grid cell in the corrected height matrix into three-dimensional points, and collect all three-dimensional points as the expected material pile point cloud; S36, calculate the arithmetic mean of the net height of the material pile of each grid cell corresponding to the corrected height matrix, and calculate the standard deviation of the expected height distribution based on the net height of the material pile of each grid cell and the arithmetic mean.

[0013] Preferably, step S4 includes the following steps: S41, the expected material pile point cloud is divided into grids to obtain the material pile surface height of each grid cell; S42, if the height of the material pile surface of any grid cell exceeds the preset safe height limit at that grid cell, then the candidate unloading point is excluded; S43, the candidate unloading sites where the surface height of the material pile in each grid cell does not exceed the corresponding upper limit of the safe height are selected as candidate unloading sites that meet the preset constraints, and the one with the smallest expected height distribution standard deviation is selected as the target unloading site.

[0014] Preferably, the step S3 is followed by the following step: Obtain the height of the material pile at a fixed point measured by at least one radar level gauge fixedly installed on the silo truss; The expected material pile point cloud is divided into grids, and the surface height of the material pile at the grid cell corresponding to the fixed point is compared with the height of the material pile at the fixed point. If the deviation between the height of the material pile surface and the height of the material pile at the fixed point exceeds a preset verification threshold, the point cloud of the current material pile is re-acquired. If the deviation does not exceed the preset verification threshold, the expected point cloud of the material pile is used in step S4.

[0015] Preferably, step S12 includes the following steps: S121, Align the geometric center of the template point cloud with the geometric center of the target point cloud to establish an initial correspondence; S122, For each point in the template point cloud, find the nearest neighbor point with the Euclidean distance in the target point cloud to form a set of corresponding point pairs; S123, based on the set of corresponding point pairs, calculate the rigid body transformation that minimizes the registration error between the template point cloud and the target point cloud, the rigid body transformation including a rotation matrix and a translation vector; S124, Apply the rotation matrix and the translation vector to the template point cloud to obtain the transformed template point cloud; S125, repeat steps S122~S124 until the registration error is less than the preset convergence threshold, and obtain the registered template point cloud.

[0016] Preferably, the slope constraint correction of the superimposed height matrix in step S34 includes the following steps: Traverse each grid cell corresponding to the superimposed height matrix to obtain the height difference between each grid cell and its adjacent grid cells and the spacing D between each grid cell and its adjacent grid cells; When the height difference is greater than D If tanα is used, then the height of the grid cell is reduced to the height of the adjacent grid cell plus D. The value of tanα, where α is the angle of repose of the material.

[0017] The present invention also provides an automatic material placement system for a sand and gravel stockpile unloading trolley, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the automatic material placement method for the sand and gravel stockpile unloading trolley as described above.

[0018] Compared with the prior art, the present invention has the following beneficial effects: (1) This invention solves the technical problem of untimely detection of material pile height and potential safety hazards in manual material placement. By acquiring the point cloud of the material pile after filtering out fixed structural points and completing the process, the global accurate acquisition of the three-dimensional shape of the material pile is achieved. By using the hard constraint that the surface height of the material pile of each grid unit does not exceed the preset safety height limit, any candidate sites that may cause local over-limits are eliminated before unloading. The global termination condition is that there are no candidate sites that meet the safety constraints. This ensures that the material placement operation automatically stops within the safety boundary, transforming the passive detection after the fact into proactive control before the fact. It realizes the search for the optimal material placement within the safety boundary and eliminates potential safety hazards such as chute blockage, increased walking pressure, and motor overload from the source.

[0019] (2) This invention solves the technical problem of insufficient uniformity of material piles caused by relying on experience to find low points and fill them locally in manual material placement. By performing virtual unloading simulation on each candidate unloading site, the shape of the material pile after unloading is predicted with the material's angle of repose as a physical constraint, and the standard deviation of the expected height distribution is calculated as a quantitative index of uniformity. Among the candidate sites that meet the safety constraints, the one that minimizes the standard deviation of the expected height distribution is selected as the target unloading site. This achieves a quantitative selection decision with the goal of optimal global uniformity, significantly improving the uniformity of the material pile while ensuring safety, minimizing the waste of storage space caused by uneven material pile height, and improving the space utilization rate of the storage silo. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0021] Figure 1 This is a schematic flowchart of one embodiment of the present invention; Figure 2 This is a schematic diagram of a movable device moving back and forth on a track to perform radar scanning, according to one embodiment of the present invention. Figure 3 This is a schematic diagram of the structure of a movable device according to an embodiment of the present invention; Figure 4 This is a schematic diagram of a target point cloud collected in an actual engineering project according to one embodiment of the present invention; Figure 5 for Figure 4 A diagram from another perspective; Figure 6 This is a schematic diagram of the point cloud of the material pile after differential filtering in one embodiment of the present invention; Figure 7 This is a schematic diagram of the structure of the silo and its internal equipment in one embodiment of the present invention; Figure 8 for Figure 7 A diagram from another perspective.

[0022] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings.

[0023] Explanation of icon numbers: 10. Movable device; 20. Track; 30. Single-line lidar; 40. Hopper; 50. Unloading trolley; 60. Truss. Detailed Implementation

[0024] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0025] 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 a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0026] Furthermore, the technical solutions of the various embodiments can be combined with each other, but only if they are feasible for those skilled in the art. If the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such combination of technical solutions does not exist and is not within the scope of protection claimed by this invention.

[0027] Please refer to Figures 1 to 8 The present invention provides an automatic material placement method for a sand and gravel stockpile unloading trolley, comprising the following steps: S1, acquire the ground reference height of the silo and the point cloud of the current material pile; wherein, the point cloud of the material pile is three-dimensional point cloud data that has been filtered out of fixed structural points of the silo and has been completed; It is worth noting that, preferably, the ground reference height of the silo is presented in an aggregate form, a two-dimensional data set associated with the horizontal position of the silo 40, providing the ground height value at any horizontal position within the silo 40. When the ground of the silo 40 is flat, the ground reference height value of each grid cell is the same and is a constant value; when the ground of the silo 40 has uneven features such as slopes or potholes, the ground height value of each grid cell may be different. Using an aggregate form ensures that the subsequently calculated net height of the material pile accurately reflects the true thickness of the material pile, rather than a false height affected by the undulations of the ground itself.

[0028] The current point cloud of the material pile can be obtained based on the moving scan of the single-line lidar 30, or based on the fixed or moving scan of the three-dimensional lidar. The following steps S11~S14 provide a preferred method for obtaining the point cloud.

[0029] Fixed structural points include point cloud data corresponding to all non-material permanent building structures within the silo (60 points), such as ground points, wall points, column points, truss points, and beam points. These fixed structural points need to be filtered out in the subsequent differential filtering stage to ensure that the resulting point cloud of the material pile represents the aggregate. Since subsequent calculations of the height distribution standard deviation and virtual unloading simulations are based on pure material pile surface data, mixing in fixed structural data—such as vertical points on walls being much higher than the material pile surface, or column points being mistakenly identified as high points in the material pile—will severely distort the uniformity index and render the spatial feasibility constraint judgment ineffective. Secondly, radar scanning is affected by column obstruction and material pile self-obstruction, resulting in blank areas in the material pile point cloud. The grid cells in these blank areas have no data, making standard deviation calculation impossible and corresponding virtual unloading simulations unavailable. Therefore, filling in these blank areas ensures the spatial continuity and completeness of the material pile point cloud.

[0030] S2, determine at least one candidate unloading site; S3, Perform a virtual unloading simulation for each candidate unloading site to generate the expected material pile point cloud corresponding to each candidate unloading site, and calculate the expected height distribution standard deviation of the expected material pile point cloud; S4. Among the candidate unloading sites that meet the preset constraints, select the one with the smallest standard deviation of the expected height distribution as the target unloading site. The preset constraints are: after the expected material pile point cloud is divided into grids, the surface height of the material pile in each grid cell does not exceed the preset safe height limit at that grid cell. S5, control the unloading trolley 50 to move to the target unloading point for unloading. When the material pile height at the target unloading point reaches the preset single-point unloading height threshold or the unloading amount reaches the preset single-time unloading amount, stop unloading at the target unloading point and return to step S1; until there are no candidate unloading points that meet the preset constraints in step S4, terminate the entire material distribution process. When the unloading trolley 50 moves on the track 20, there is a suitable height between the bottom surface of the chute and the ground of the hopper 40. When the material pile height is too high, the chute will be blocked by the material pile, leading to blockage or equipment damage. Therefore, the single-point unloading height threshold must be less than the maximum allowable material pile height at that location. For example, the single-point unloading height threshold can be set to a certain percentage of the upper limit of the safe height, such as 80% to 90%, to leave a safety margin, and the single-point unloading height threshold must not exceed the upper limit of the safe height (e.g., ...). Figure 7(Safety height limits in the diagram). In actual production, operators can estimate the average unloading amount per round of material placement based on the expected total volume of the stockpile and the planned number of placement rounds. The unloading amount per round is usually taken as an appropriate proportion of the average unloading amount to ensure the controllability of the material placement process. Those skilled in the art can set this as needed.

[0031] The unloading trolley 50 is controlled to move to the target unloading point for unloading. During the unloading process, the height of the material pile at the point can be monitored in real time by the vehicle-mounted radar level gauge installed on the chute of the unloading trolley 50. When the height of the material pile at the target unloading point reaches the preset single-point unloading height threshold, or the cumulative unloading amount reaches the preset single unloading amount, the unloading at the target unloading point is stopped.

[0032] After completing this unloading, return to step S1 to reacquire the point cloud of the current material pile, and execute steps S2 to S4 to determine the next target unloading site and execute unloading. This process is repeated until, during a certain execution of step S4, all candidate unloading sites fail the spatial feasibility constraint check after the virtual unloading simulation, meaning there are no candidate unloading sites that meet the preset constraint conditions. At this point, there are no unloading sites within the silo 40 that can safely accommodate the new material, and the entire material placement process is terminated.

[0033] This invention solves the technical problem of untimely detection of material pile height and potential safety hazards in manual material placement. By acquiring a point cloud of the material pile after filtering out fixed structural points and completing the data, a global and accurate acquisition of the three-dimensional shape of the material pile is achieved. By using the constraint that the surface height of the material pile in each grid unit does not exceed a preset safety height limit as a hard constraint, any candidate sites that may cause local exceedances are eliminated before unloading. The global termination condition is that there are no candidate sites that meet the safety constraints, ensuring that the material placement operation automatically stops within the safety boundary. This transforms the passive detection of material piles after the fact into proactive control before the fact, realizing the search for the optimal material placement within the safety boundary and eliminating safety hazards such as chute blockage, increased walking pressure, and motor overload from the source.

[0034] This invention solves the technical problem of insufficient uniformity in material piles caused by relying on experience to find low points and fill them locally during manual material placement. By performing virtual unloading simulations on each candidate unloading site, the post-unloading pile morphology is predicted using the material's angle of repose as a physical constraint, and the standard deviation of the expected height distribution is calculated as a quantitative indicator of uniformity. Among the candidate sites that meet safety constraints, the one that minimizes the expected height distribution standard deviation is selected as the target unloading site. This achieves a quantitative optimization decision with the goal of optimal global uniformity, significantly improving the uniformity of the material pile while ensuring safety, minimizing storage space waste caused by uneven material pile heights, and increasing the space utilization rate of the storage silo by 40%.

[0035] It is also worth noting that in uneven material piles, local high points can cause the system to terminate material distribution prematurely, while many low-lying areas still need to be filled. The unloading trolley 50 then has to frequently turn back and forth within the hopper 40 to find low points, resulting in long empty runs and low operational efficiency. Therefore, improving the uniformity of the material pile, allowing the height of each area to increase synchronously, enables the unloading trolley 50 to operate continuously within the length of the hopper 40, reducing ineffective backtracking and increasing the proportion of effective material distribution time.

[0036] In a preferred embodiment, step S1 includes the following steps: S11, acquire the template point cloud collected in the empty warehouse state, and filter the ground point cloud from the template point cloud according to the height threshold, and use it as the ground reference height of the warehouse. The template point cloud is a three-dimensional point cloud data obtained by using a point cloud scanning device to perform a global scan of the silo 40 when no materials are piled up in the silo 40. The template point cloud contains complete fixed structural information of the silo 40, including fixed structures such as the ground, walls, and columns, which are the things that are already in the silo 40.

[0037] In a preferred embodiment, the movable device 10 equipped with a single-line lidar 30 can be controlled to move along the track 20; wherein the single-line lidar 30 is vertically mounted downwards at the bottom of the movable device 10; as shown Figure 3 As shown, the movable device 10 in this embodiment uses a track trolley, which is installed on a fixed track 20 above the silo 40. The track 20 is located on the bottom side of the truss 60, offset from the unloading area of ​​the unloading trolley 50, and is laid along the length of the silo 40 (i.e., the front-to-back direction of the material pile). The track trolley can be driven by a servo motor. A single-line laser radar 30 is fixed vertically downward at the bottom of the track trolley, and the scanning plane is perpendicular to the movement direction of the track trolley. A schematic diagram of the radar scanning performed by the track trolley moving back and forth is shown below. Figure 2 As shown, when the single-line lidar 30 rotates one revolution or within a preset angle range in a vertical plane, it can acquire a two-dimensional cross-sectional point cloud of the area where the material pile is located. In other embodiments, the two-dimensional cross-sectional point cloud can also be obtained using other existing forms of mobile platforms.

[0038] During the movement, the single-line lidar 30 is triggered to scan according to a preset time interval or a preset distance interval. Each time the single-line lidar 30 is triggered, it rotates one revolution in the vertical plane or within a preset angle range and acquires a frame of two-dimensional cross-sectional point cloud. The system sends a trigger signal to the radar every 0.5 seconds according to a preset time interval, for example, a preset time interval of 0.5 seconds. After receiving the signal, the radar acquires a frame of two-dimensional cross-sectional point cloud in polar coordinate form containing multiple points. The system acquires the position information of the movable device 10 in the direction of movement and associates the position information with the two-dimensional cross-sectional point cloud collected at the corresponding time. When the radar completes a frame scan and outputs data, the system packages the position information with the radar data frame to form a position-cross-section data pair for storage.

[0039] For each frame of the 2D cross-sectional point cloud, the polar coordinates are first converted to local Cartesian coordinates centered on the single-line lidar 30. Then, based on the vehicle's position, the local coordinates are transformed to the global coordinate system (e.g., ...). Figure 2 As shown, the X-axis is perpendicular to track 20, the Y-axis is perpendicular downwards, and the Z-axis extends along the direction of track 20.

[0040] All the converted 2D cross-sectional point clouds are stacked in ascending order of Z-coordinate to form a complete 3D point cloud. The point cloud collected in the empty material pile state of the silo is called the template point cloud (including fixed structures such as the ground, walls, and pillars); the point cloud collected in the material pile state is called the target point cloud (including fixed structures and the material pile), such as... Figures 4 to 5 As shown.

[0041] It is important to note that Figures 4 to 6 The point cloud image shows an uneven material pile shape. This is because the bottom of the hopper 40 has multiple discharge holes. After the unloading trolley 50 completes the material distribution, the bottom discharge causes a change in the shape of the material pile, resulting in a highly uneven and irregular shape. This shape is caused by the downstream discharge process after the material distribution is completed, and it belongs to a different operation stage from the material pile surface uniformity pursued in the material distribution stage of this application. The two do not conflict.

[0042] It is worth noting that the point cloud with the lowest height value in the template point cloud (e.g., the 5% of point clouds with the lowest height value) is selected, filtered to remove outlier noise points, and then used as the ground reference point cloud. This overcomes the problem that operators cannot accurately know the actual ground shape of the silo 40 during manual material placement. The ground reference point cloud provides accurate ground elevation data at each horizontal position for subsequent calculation of the net height of the material pile.

[0043] S12, acquire the target point cloud collected under the current material pile state, perform spatial transformation on the template point cloud to align the template point cloud with the target point cloud, and obtain the registered template point cloud; This implementation achieves spatial alignment between the template point cloud and the target point cloud through spatial registration, providing a computational basis for subsequent differential filtering.

[0044] Further, step S12 includes the following steps: S121, Align the geometric center of the template point cloud with the geometric center of the target point cloud to establish an initial correspondence; S122, for each point in the template point cloud, find the nearest neighbor in the target point cloud with the smallest Euclidean distance to form a corresponding point pair set; by traversing every 3D point in the template point cloud, calculate the Euclidean distance between the point and all points in the target point cloud, and select the point in the target point cloud with the smallest Euclidean distance as the nearest neighbor of the template point. Each pair of template points and its nearest neighbor forms a corresponding point pair, and the set of all corresponding point pairs is the corresponding point pair set for the current iteration.

[0045] S123, based on the set of corresponding point pairs, calculate the rigid body transformation that minimizes the registration error between the template point cloud and the target point cloud, the rigid body transformation including a rotation matrix and a translation vector; S124, apply the rotation matrix and the translation vector to the template point cloud to obtain the transformed template point cloud; calculate the centroid of the corresponding point set of the template point cloud and the centroid of the corresponding point set of the target point cloud respectively; subtract the centroid of each of the two corresponding point sets to obtain centroid-free point sets; construct the covariance matrix of the centroid-free point sets; perform singular value decomposition on the covariance matrix, and calculate the rotation matrix and translation vector from the decomposition results. Singular value decomposition is a standard numerical method for solving optimal rigid body transformations, and mature techniques will not be elaborated here.

[0046] S125, repeat steps S122-S124 until the registration error is less than a preset convergence threshold, obtaining the registered template point cloud. The registration error is the root mean square or average value of the Euclidean distance between all corresponding point pairs in the current iteration. After each step S124, calculate the registration error for the current iteration. If the registration error is less than the preset convergence threshold (e.g., 1 mm), it indicates that the spatial deviation between the template point cloud and the target point cloud has been reduced to an acceptable range. Stop the iteration and use the currently transformed template point cloud as the final registered template point cloud.

[0047] S13, for each point in the target point cloud, find the nearest neighbor of the point in the registered template point cloud, calculate the Euclidean distance between the point and the nearest neighbor, retain the points with Euclidean distance greater than the preset distance threshold as material pile points, and filter out the remaining points as fixed structure points to obtain the preliminary material pile point cloud. Specifically, a spatial index can be built for the registered template point cloud; After spatial alignment of the template point cloud and the target point cloud, the registered template point cloud is obtained. To efficiently query the nearest neighbor of each point in the target point cloud in the template point cloud, a KDTree data structure is used to construct a spatial index for the template point cloud. KDTree is a binary tree that partitions and stores points in k-dimensional space; mature algorithms will not be elaborated upon here.

[0048] Calculate the nearest distance from each point in the target point cloud to the registered template point cloud; For each point in the target point cloud, a nearest neighbor search is performed using the KDTree spatial index to find the point in the registered template point cloud that has the closest spatial Euclidean distance to the target point, and the Euclidean distance is calculated. Points whose nearest distance is greater than a preset filtering threshold are retained, and points whose nearest distance is less than or equal to the preset filtering threshold are filtered out, resulting in a filtered point cloud, as shown below. Figure 6 As shown, the closer the color is to purple, the higher the altitude; the closer the color is to yellow, the lower the altitude.

[0049] The preset filtering threshold can be an empirical parameter used to distinguish between fixed structure points and material pile points. As a preferred example, a preset filtering threshold of 5cm is used. For each target point, if the nearest Euclidean distance is greater than 5cm, it is identified as a material pile point and retained; if the nearest Euclidean distance is less than or equal to 5cm, it is identified as a fixed structure point and filtered out. Through the above steps, points in the target point cloud that coincide with or are close to the template point cloud in spatial position can be removed, i.e., all fixed structure points, while material pile surface points are retained, resulting in a filtered point cloud.

[0050] S14, Identify the void regions in the preliminary material pile point cloud, and fill and complete the void regions using a constraint interpolation method based on the physical properties of the material to obtain the material pile point cloud.

[0051] In a preferred embodiment, S14 includes the following steps: S141, Identify the hole regions in the filtered point cloud, and divide the hole regions into grids to obtain several grid points; In the filtered point cloud, because the point cloud is discrete, void regions appear as continuous blank areas with point cloud density significantly lower than the average density. During radar scanning, these voids can be caused by obstructions from buildings such as pillars, as well as self-obstruction from the material pile itself.

[0052] For identified void regions, a voxelization statistical method can be used: the point cloud is divided into regular grids on a horizontal plane, and the number of points in each grid is counted. If a grid has 0 points, it is marked as a candidate void unit; adjacent candidate void units are merged to form void regions.

[0053] The void region is divided into grids on a horizontal plane. The grid side length can be set according to the radar sampling interval, preferably 1.0 to 1.5 times the sampling interval. After gridding, the vertex of each grid cell (i.e., the intersection of grid lines) is the grid point. The height of the grid point is unknown and needs to be determined by subsequent interpolation.

[0054] S142, Based on the known points around the cavity region, bilinear interpolation is used to determine the initial height of each grid point; wherein, the known points are points located on the boundary of the cavity region in the filtered point cloud; The known points around the cavity region are those points in the filtered point cloud located on the boundary of the cavity region (or immediately adjacent to the cavity). Specifically, for the boundary grid cell of the cavity region, the point cloud points in the non-cavity grid cell are considered as known points, or the point cloud points on the boundary of the cavity region can be directly taken. The known points provide the true height information of the cavity edge and are the basis for interpolation. For each grid point within the cavity region, first find four known points (or grid points) around the cavity as interpolation nodes. These four points should be located in the upper left, upper right, lower left, and lower right quadrants of the grid point, forming a rectangle surrounding the grid point. The initial height of the grid point is obtained through bilinear interpolation.

[0055] S143, calculate the slope between each grid point and the known neighboring points. When the slope exceeds the preset angle of repose threshold, correct the height of the grid point so that the corrected slope is not greater than the preset angle of repose threshold. The known neighboring points include known points on the boundary of the void region (original point cloud points) and other grid points that have been corrected before this iteration (i.e., corrected grid points). The neighborhood range is usually a circular or square area centered on the grid point with a radius of 1.5 times the grid side length. The slope between each grid point and the known neighboring points is calculated. In the natural accumulation of sand and gravel, the slope will not exceed the angle of repose of the material (e.g., the angle of repose of sand and gravel is approximately 30°~40°). By calculating the slope, the conformity of the interpolation result to physical laws can be quantified. The preset angle of repose threshold is set according to the material type; for example, for sand and gravel, the angle of repose can be 40°. The specific method for correcting the height of the grid point is to correct the height to a new value that is lower than the original height and satisfies the slope constraint. If the slope between a grid point and the known neighboring points exceeds the preset angle of repose threshold, then the height of the grid point is reduced to a height value that makes the slope between them equal to the preset angle of repose threshold, based on the height of the known neighboring points.

[0056] S144, iteratively execute step S143, wherein the known points in the neighborhood include known points on the boundary of the void region and the corrected grid points, until the slope of all grid points is not greater than the preset angle of repose threshold, and obtain the completed point cloud.

[0057] Correcting a grid point may affect the slope of other grid points in its neighborhood. Step S143 is executed iteratively. In each iteration, all grid points within the void region are traversed, and the slopes are calculated and corrected according to step S143. The iteration stops when the slope of all grid points in a given iteration is no greater than the angle of repose threshold. After the iteration is complete, the height values ​​of all grid points are determined. The grid points are then converted to a 3D point cloud format, merged with the original filtered point cloud, and duplicate points are removed to obtain the completed point cloud, i.e., the stockpile point cloud.

[0058] The point cloud of the material pile obtained through the above steps is a three-dimensional point cloud data after filtering out the fixed structural points of the silo and completing the data. The point cloud of the material pile accurately reflects the complete three-dimensional surface morphology of the current material pile within the silo 40, providing an accurate data basis for subsequent material placement decisions.

[0059] In a preferred embodiment, step S2 includes the following steps: S21, extract the cross-sectional height curve from the material pile point cloud along the moving direction of the unloading trolley 50; wherein, the cross-sectional height curve reflects the highest height of the material pile surface at each sampling point along the moving direction, and the sampling point is the position point divided along the moving direction of the unloading trolley 50 by a preset sampling interval. In a preferred embodiment, the step S21 of extracting the cross-sectional height curve along the moving direction of the unloading trolley 50 includes the following steps: S211, along the moving direction of the unloading trolley 50, determine each sampling position point according to the preset sampling interval; taking the starting point of the unloading trolley 50's moving path as the first sampling position point, take a sampling position point every preset sampling interval along the moving direction until the entire moving path is covered. The preset sampling interval can be set according to the point cloud density and material placement accuracy requirements, for example, 0.5m. The smaller the sampling interval, the higher the spatial resolution of the candidate sites, but the computational load increases accordingly; the larger the sampling interval, the higher the computational efficiency, but local low points may be missed.

[0060] S212, for each sampling location, the maximum height of all points within a cross-section perpendicular to the direction of movement in the point cloud of the material pile at that sampling location is taken as the cross-sectional height corresponding to that sampling location. It should be noted that because the point cloud itself is discretely sampled, there may not be enough data points near the plane for stable height statistics. To obtain statistically significant material pile surface height information at the sampling location, a preset distance (i.e., bandwidth), for example ±0.25m, is extended in both directions before and after the sampling location to collect all material pile point cloud points falling within this strip area. Then, the maximum height among these points is taken as the cross-sectional height corresponding to that sampling location, and the highest point within the cross-section best represents the top shape of the material pile at that location.

[0061] S213, the cross-sectional heights corresponding to each sampling location point are connected sequentially according to the moving direction to form the cross-sectional height curve. Connecting them sequentially in ascending order of the moving direction forms a continuous, smooth curve.

[0062] S22, the sampling points on the cross-sectional height curve whose height is lower than the average height of the entire cross-sectional height curve and whose height is lower than the immediately preceding and following sampling points are determined as the candidate unloading points.

[0063] The arithmetic mean height of the entire cross-sectional height curve is obtained by summing the cross-sectional height values ​​corresponding to all sampling points on the curve and dividing by the total number of sampling points. Along the direction of movement, the height of each sampling point is compared with the height of its immediate preceding and following sampling points. If the height of a sampling point is lower than both its preceding and following sampling points, then that sampling point is a local minimum on the cross-sectional height curve. These points are then selected as the lowest points in the overall low-lying area and should be prioritized for filling during the fabric laying operation.

[0064] This step uses the arithmetic mean height of the entire curve as a reference benchmark, enabling the screening of candidate sites to be quantified and automated. This overcomes the arbitrariness and inaccuracy of operators relying on subjective experience to judge where the low points are in manual fabric laying, and eliminates the subjective differences in manual judgment.

[0065] In a preferred embodiment, step S3 includes the following steps: S31, the material pile point cloud is divided into horizontal grids to obtain several grid cells. The maximum height of each material pile point within each grid cell is taken as the surface height of the material pile for that grid cell. The net height of the material pile is obtained by subtracting the reference height of the silo ground at that grid cell from the surface height of the material pile. The material pile point cloud is projected onto a horizontal plane and divided into regular rectangular grids with preset side lengths. The grid side length can be determined according to the point cloud sampling interval, usually 1.0 to 1.5 times the sampling interval, to ensure that each grid contains a sufficient number of point cloud points for stable height statistics. For each grid cell, all material pile point cloud points within the grid cell are traversed, and the maximum height of each point is taken as the surface height of the material pile for that grid cell. The surface height of the material pile should reflect the upper surface contour of the material pile at that location, representing the true top shape of the material pile at that location.

[0066] S32, with the candidate unloading site as the center, construct an accumulation body according to the preset single unloading amount and material repose angle, and determine the accumulation height of the accumulation body at each grid cell; For each candidate unloading point, an accumulation body is constructed centered on the projected coordinates of the candidate unloading point on the horizontal plane, based on the preset single unloading volume and material repose angle constraints. Specifically, the accumulation body is preferably conical. It is worth noting that the chute of the unloading trolley 50 is usually circular in cross-section. When the material falls freely from the chute opening to the surface of the stockpile, under the combined action of gravity, inter-particle friction, and friction between particles and the stockpile surface, the material will naturally slide outwards from the point of impact, eventually forming a stable conical accumulation body whose generatrix makes an angle with the horizontal plane equal to the material repose angle. The material repose angle is the maximum angle formed between the free surface of the granular material and the horizontal plane in its natural accumulation state, and is an inherent physical property parameter of the material. Therefore, under stable conditions, the shape of the accumulation body formed by single-point unloading is close to a standard cone, and this geometric approximation is valid within engineering accuracy.

[0067] Its base radius R and height H are determined by the preset single unloading amount Q and the material repose angle α through the following geometric relationship: Formula 1 is thus obtained: ; Formula 2: ; Combining formulas one and two, we get: ; .

[0068] Therefore, for each grid cell, the horizontal distance from the center point of the grid cell to the center of the candidate unloading point is calculated. The stacking height of the grid cell is determined based on the height distribution of the stack at different horizontal distances from the center of the candidate unloading point. If a grid cell is located outside the bottom surface of the stack, the stacking height of the grid cell is zero.

[0069] For a conical accumulation, the accumulation height h at any horizontal distance d from the center is determined by the following formula: if d ≤ R, then h = H × (1 - d / R); if d > R, then h = 0; where R is the radius of the cone's base and H is the cone's height. Therefore, for each grid cell, the horizontal distance from the center of the grid cell to the center of the candidate unloading point is calculated, and the accumulation height at that grid cell is determined by substituting this distance into the above formula. If the grid cell is located outside the bottom surface of the accumulation, the accumulation height of that grid cell is zero.

[0070] S33, for each grid cell, take the larger value between the stacking height and the surface height of the material pile as the height of the superimposed grid cell, and form a superimposed height matrix; For each grid cell, the stacking height value determined in step S32 is compared with the material pile surface height value determined in step S31. The larger of the two values ​​is taken as the height of the superimposed grid cell. The superimposed heights of all grid cells are arranged according to their grid positions to form a superimposed height matrix. Newly unloaded material covers the surface of the original material pile. If the new material is higher than the old material, the surface is raised; if the new material is lower than the old material, the position remains unchanged.

[0071] S34, perform slope constraint correction on the superimposed height matrix, reduce the part of the local slope that exceeds the material repose angle to meet the limit value of the material repose angle, and obtain the corrected height matrix; Iterate through each grid cell corresponding to the superimposed height matrix. For each grid cell, calculate the slope between that grid cell and its adjacent grid cells. The slope between a given grid cell and its adjacent grid cells is defined as the absolute value of their height difference divided by their center distance. If the slope between a given grid cell and any adjacent grid cell exceeds the tangent of the material's angle of repose, i.e., the height difference between the given grid cell and its adjacent grid cells is greater than the distance D between them... The product of these two values ​​indicates that the surface of the stacked material pile is too steep, preventing the material from settling stably and causing it to slide. In this case, the height of the grid cell should be reduced so that the slope at that location is exactly equal to... The value is corrected to the height of adjacent grid cells plus the grid spacing. The product of.

[0072] Furthermore, the slope constraint correction of the superimposed height matrix in step S34 includes the following steps: Traverse each grid cell corresponding to the superimposed height matrix to obtain the height difference between each grid cell and its adjacent grid cells and the spacing D between each grid cell and its adjacent grid cells; When the height difference is greater than D When tanα is reached, the height of the grid cell is reduced to the height of the adjacent grid cell plus D. The value of tanα, where α is the angle of repose of the material.

[0073] S35, convert the horizontal position coordinates and corresponding corrected heights of each grid cell in the corrected height matrix into three-dimensional points, and collect all three-dimensional points as the expected material pile point cloud; Traverse all grid cells in the corrected height matrix. For each grid cell, generate a 3D point using the horizontal coordinates of its center point and the corrected height of that grid cell. The 3D points obtained from all grid cells constitute the expected stockpile point cloud.

[0074] S36, calculate the arithmetic mean of the net height of the material pile of each grid cell corresponding to the corrected height matrix, and calculate the standard deviation of the expected height distribution based on the net height of the material pile of each grid cell and the arithmetic mean.

[0075] The arithmetic mean μ of the net height of the material pile in all effective grid cells of the corrected height matrix is ​​calculated. The net height of the material pile in each grid cell corresponding to the corrected height matrix is ​​the corrected height of that grid cell minus the reference height of the silo ground at that grid cell. The sum of squares of the deviations between the net height of the material pile in each grid cell and the arithmetic mean is calculated. The sum of squares of the deviations is divided by the number of effective grid cells N to obtain the variance. The square root of the variance is taken to obtain the standard deviation of the expected height distribution. The standard deviation of the expected height distribution is an index that quantifies the global uniformity of the material pile after unloading at the candidate unloading site. The smaller the standard deviation of the expected height distribution, the more concentrated, uniform and flat the expected height distribution of the material pile surface.

[0076] In a preferred embodiment, step S4 includes the following steps: S41, the expected material pile point cloud is divided into grids to obtain the material pile surface height of each grid cell; for the expected material pile point cloud corresponding to each candidate unloading point, the same grid division operation as in step S31 is performed, dividing the expected material pile point cloud into regular rectangular grids under the same horizontal coordinate system, using the same grid origin and the same preset side length. For each grid cell, all three-dimensional points of the expected material pile point cloud within the grid are traversed, and the maximum value of the height of each point is taken as the material pile surface height of the grid cell.

[0077] S42, if the height of the material pile surface of any grid cell exceeds the preset safe height limit at that grid cell, then the candidate unloading point is excluded; For each candidate unloading point, all grid cells generated in step S41 are traversed. For each grid cell, the height of the material pile surface is compared with the preset upper limit of the safe height at that grid cell. The upper limit of the safe height is the maximum allowable height value preset for each grid cell, which can be determined based on the minimum safe clearance of the unloading trolley 50 chute at each horizontal position of the silo 40. For example, the upper limit of the safe height = the elevation of the moving track 20 of the unloading trolley 50 at that location - the minimum safe clearance of the chute. The minimum safe clearance of the chute refers to the minimum safe distance that must be maintained between the bottom of the unloading chute and the highest point of the material pile. It is a fixed empirical threshold (e.g., 50cm) set according to the equipment safety specifications. Its function is to prevent the chute from colliding with the material pile and avoid material blockage or equipment damage. The upper limit of the safe height is determined based on the spatial distribution of the top structure of the silo 40. If there is a beam, the height is lowered; if there is no structure, the normal value is used.

[0078] If the height of the material pile surface in a certain grid cell exceeds the preset safety height limit, it means that unloading at this candidate site will cause at least one location of the material pile to exceed the safety allowable range, potentially leading to safety accidents such as chute blockage or equipment collision. In this case, the candidate unloading site is immediately determined to not meet the preset constraints and is excluded from the candidate set, no longer participating in the subsequent selection comparison.

[0079] S43, the candidate unloading sites where the surface height of the material pile in each grid cell does not exceed the corresponding upper limit of the safe height are selected as candidate unloading sites that meet the preset constraints, and the one with the smallest expected height distribution standard deviation is selected as the target unloading site.

[0080] All candidate unloading sites that pass the spatial feasibility check and are not excluded are identified as candidate unloading sites that meet the preset constraints. The standard deviation of the expected height distribution of each of these candidate unloading sites that meet the constraints is compared, and the candidate unloading site with the smallest standard deviation of expected height distribution is selected as the target unloading site.

[0081] This implementation method achieves optimal decision-making for the overall uniformity of the material pile while ensuring operational safety. It overcomes the shortcomings of relying solely on experience to determine the unloading location and failing to quantify and compare the impact of different locations on safety and uniformity. This method achieves the technical effect of improving uniformity while ensuring safety.

[0082] Furthermore, it also includes the following steps: Obtain the height of the material pile at a fixed point measured by at least one radar level gauge fixedly installed on the silo truss 60; Specifically, at least one fixed radar level gauge is installed at a fixed position on the traveling truss 60 of the unloading trolley 50. The fixed radar level gauges are installed at different spatial coordinates on the truss 60, for example, at one-quarter, one-half, and three-quarters of the span of the truss 60, respectively, to cover multiple representative points in the silo 40 space. These fixed radar level gauges are stationary and do not move with the unloading trolley 50. The coordinates of their installation positions in the global coordinate system are pre-calibrated, and the fixed radar level gauges can accurately measure the height of the material pile surface below their respective fixed points.

[0083] The expected material pile point cloud is divided into grids, and the surface height of the material pile at the grid cell corresponding to the fixed point is compared with the height of the material pile at the fixed point. After each update of the current stockpile point cloud, the same meshing operation as in step S31 is performed on the expected stockpile point cloud. In the meshed expected stockpile point cloud, the mesh cells corresponding to the installation positions of each fixed radar level gauge are located. The extracted stockpile surface height value of the mesh cell is compared with the measured stockpile height value of the corresponding fixed radar level gauge, and the absolute deviation between the two is calculated. If there are multiple fixed radar level gauges, the absolute deviation at each fixed point is calculated separately.

[0084] If the deviation between the height of the material pile surface and the height of the material pile at the fixed point exceeds a preset verification threshold, the point cloud of the current material pile is re-acquired. If the deviation does not exceed the preset verification threshold, the expected point cloud of the material pile is used in step S4.

[0085] The calculated absolute deviations at each fixed point are compared with a preset verification threshold, which is an engineering parameter, for example, 5 cm. If the absolute deviations at each fixed point do not exceed the preset verification threshold, it indicates that the expected material pile point cloud constructed based on the point cloud is highly consistent with the independent measurement values ​​of the radar level gauge, the model is realistic and reliable, and can be used for subsequent material placement decisions.

[0086] The present invention also provides an automatic material placement system for a sand and gravel stockpile unloading trolley, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the automatic material placement method for the sand and gravel stockpile unloading trolley as described above.

[0087] like Figures 7 to 8 As shown in the figure, this embodiment illustrates the structure of the silo 40 and its internal equipment. The unloading trolley 50 is movably mounted on the truss 60. The bottom of the silo 40 is provided with a discharge port. The feeding and discharging processes are upstream and downstream processes.

[0088] The above are merely preferred embodiments of the present invention and do not limit the scope of protection of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention’s specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. An automatic material placement method for a sand and gravel stockpile unloading trolley, characterized in that, Includes the following steps: S1, acquire the ground reference height of the silo and the point cloud of the current material pile; wherein, the point cloud of the material pile is three-dimensional point cloud data that has been filtered out of fixed structural points of the silo and has been completed; S2, determine at least one candidate unloading site; S3, Perform a virtual unloading simulation for each candidate unloading site to generate the expected material pile point cloud corresponding to each candidate unloading site, and calculate the expected height distribution standard deviation of the expected material pile point cloud; S4. Among the candidate unloading sites that meet the preset constraints, select the one with the smallest standard deviation of the expected height distribution as the target unloading site. The preset constraints are: after the expected material pile point cloud is divided into grids, the surface height of the material pile in each grid cell does not exceed the preset safe height limit at that grid cell. S5, control the unloading trolley to move to the target unloading point for unloading. When the height of the material pile at the target unloading point reaches the preset single-point unloading height threshold or the unloading amount reaches the preset single unloading amount, stop the unloading at the target unloading point and return to step S1. Until there are no candidate unloading points that meet the preset constraints in step S4, terminate the entire material laying process.

2. The automatic material placement method for the sand and gravel stockpile unloading trolley according to claim 1, characterized in that, Step S1 includes the following steps: S11, acquire the template point cloud collected in the empty warehouse state, and filter the ground point cloud from the template point cloud according to the height threshold, and use it as the ground reference height of the warehouse. S12, acquire the target point cloud collected under the current material pile state, perform spatial transformation on the template point cloud to align the template point cloud with the target point cloud, and obtain the registered template point cloud; S13, for each point in the target point cloud, find the nearest neighbor of the point in the registered template point cloud, calculate the Euclidean distance between the point and the nearest neighbor, retain the points with Euclidean distance greater than the preset distance threshold as material pile points, and filter out the remaining points as fixed structure points to obtain the preliminary material pile point cloud. S14, Identify the void regions in the preliminary material pile point cloud, and fill and complete the void regions using a constraint interpolation method based on the physical properties of the material to obtain the material pile point cloud.

3. The automatic material placement method for the sand and gravel stockpile unloading trolley according to claim 1, characterized in that, Step S2 includes the following steps: S21, extract the cross-sectional height curve from the point cloud of the material pile along the moving direction of the unloading trolley; wherein, the cross-sectional height curve reflects the highest height of the material pile surface at each sampling point along the moving direction, and the sampling point is a point divided along the moving direction of the unloading trolley by a preset sampling interval; S22, the sampling points on the cross-sectional height curve whose height is lower than the average height of the entire cross-sectional height curve and whose height is lower than the immediately preceding and following sampling points are determined as the candidate unloading points.

4. The automatic material placement method for the sand and gravel stockpile unloading trolley according to claim 3, characterized in that, Step S21, extracting the cross-sectional height curve along the moving direction of the unloading trolley, includes the following steps: S211, along the moving direction of the unloading trolley, determine each sampling position point according to the preset sampling interval; S212, For each sampling location point, take the maximum value of the height of each point in the cross section perpendicular to the moving direction of the material pile point cloud at that sampling location point, and take it as the cross section height corresponding to that sampling location point; S213, the cross-sectional heights corresponding to each sampling location point are connected sequentially according to the moving direction to form the cross-sectional height curve.

5. The automatic material placement method for the sand and gravel stockpile unloading trolley according to claim 3, characterized in that, Step S3 includes the following steps: S31, divide the material pile point cloud into horizontal plane grids to obtain several grid units. Take the maximum height of each material pile point in each grid unit as the material pile surface height of the grid unit. Subtract the material silo ground reference height at the grid unit from the material pile surface height to obtain the net height of the material pile of the grid unit. S32, with the candidate unloading site as the center, construct an accumulation body according to the preset single unloading amount and material repose angle, and determine the accumulation height of the accumulation body at each grid cell; S33, for each grid cell, take the larger value between the stacking height and the surface height of the material pile as the height of the superimposed grid cell, and form a superimposed height matrix; S34, perform slope constraint correction on the superimposed height matrix, reduce the part of the local slope that exceeds the material repose angle to meet the limit value of the material repose angle, and obtain the corrected height matrix; S35, convert the horizontal position coordinates and corresponding corrected heights of each grid cell in the corrected height matrix into three-dimensional points, and collect all three-dimensional points as the expected material pile point cloud; S36, calculate the arithmetic mean of the net height of the material pile of each grid cell corresponding to the corrected height matrix, and calculate the standard deviation of the expected height distribution based on the net height of the material pile of each grid cell and the arithmetic mean.

6. The automatic material placement method for the sand and gravel stockpile unloading trolley according to claim 5, characterized in that, Step S4 includes the following steps: S41, the expected material pile point cloud is divided into grids to obtain the material pile surface height of each grid cell; S42, if the height of the material pile surface of any grid cell exceeds the preset safe height limit at that grid cell, then the candidate unloading point is excluded; S43, the candidate unloading sites where the surface height of the material pile in each grid cell does not exceed the corresponding upper limit of the safe height are selected as candidate unloading sites that meet the preset constraints, and the one with the smallest expected height distribution standard deviation is selected as the target unloading site.

7. The automatic material placement method for the sand and gravel stockpile unloading trolley according to claim 6, characterized in that, Step S3 is followed by the following steps: Obtain the height of the material pile at a fixed point measured by at least one radar level gauge fixedly installed on the silo truss; The expected material pile point cloud is divided into grids, and the surface height of the material pile at the grid cell corresponding to the fixed point is compared with the height of the material pile at the fixed point. If the deviation between the height of the material pile surface and the height of the material pile at the fixed point exceeds a preset verification threshold, the point cloud of the current material pile is re-acquired. If the deviation does not exceed the preset verification threshold, the expected point cloud of the material pile is used in step S4.

8. The automatic material placement method for the sand and gravel stockpile unloading trolley according to claim 2, characterized in that, Step S12 includes the following steps: S121, Align the geometric center of the template point cloud with the geometric center of the target point cloud to establish an initial correspondence; S122, For each point in the template point cloud, find the nearest neighbor point with the Euclidean distance in the target point cloud to form a set of corresponding point pairs; S123, based on the set of corresponding point pairs, calculate the rigid body transformation that minimizes the registration error between the template point cloud and the target point cloud, the rigid body transformation including a rotation matrix and a translation vector; S124, Apply the rotation matrix and the translation vector to the template point cloud to obtain the transformed template point cloud; S125, repeat steps S122~S124 until the registration error is less than the preset convergence threshold, and obtain the registered template point cloud.

9. The automatic material placement method for the sand and gravel stockpile unloading trolley according to claim 5, characterized in that, Step S34, which involves correcting the slope constraint of the superimposed height matrix, includes the following steps: Traverse each grid cell corresponding to the superimposed height matrix to obtain the height difference between each grid cell and its adjacent grid cells and the spacing D between each grid cell and its adjacent grid cells; When the height difference is greater than D If tanα is used, then the height of the grid cell is reduced to the height of the adjacent grid cell plus D. The value of tanα, where α is the angle of repose of the material.

10. An automatic material distribution system for a sand and gravel stockpile unloading trolley, characterized in that, The system includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the automatic material placement method for the sand and gravel stockpile unloading trolley as described in any one of claims 1 to 9.

Images