Full-automatic material pile measuring method

Abstract

The invention discloses a full-automatic material pile measurement method, and relates to the technical field of material pile measurement. The method comprises the following steps: firstly, deploying a plurality of measurement points on a ceiling of a stock yard, installing a radar with a holder at each measurement point, and simultaneously sending instantaneous radar measurement data and yaw and pitching data of the holder to a data processing computer through the Ethernet by the radars and the holders; determining the conversion process of the feedback data of the radar and the holder and the coordinate system, converting the single measurement value of the radar into the homogeneous coordinate of the surface point of the material pile in the system coordinate system, converging the scanning curve to form local point clouds, and splicing the local point clouds of the adjacent measurement points to obtain the overall point cloud of the material pile; on the basis, an elevation model DEM of the material pile is obtained based on the regular grid omega and the elevation of the grid points; and after hole filling and edge smoothing, calculating the volume and the weight of the material pile by using a DEM (Digital Elevation Model). According to the invention, automatic data acquisition and point cloud processing are realized, and the measurement precision and efficiency can be obviously improved.

Description

[0001] This invention is a divisional application of a patent application entitled "Fully Automatic Method for Measuring Large Stockpiles", the original application was filed on September 19, 2022, application number 202211138981.7. Technical Field

[0002] This invention relates to the field of stockpile measurement technology, and specifically to a fully automated stockpile measurement method. Background Technology

[0003] Mineral resources such as coal and ore are mostly stored in large bulk stockpiles at ports, mines, or related enterprise yards. Due to the high prices and dynamic changes in stock levels of these resources, regular checks on the volume and weight of the stockpiles are necessary. Currently, the measurement of large stockpiles primarily employs lidar-based methods, mainly divided into mobile and fixed modes. The mobile mode requires the radar to be mounted on the top of a crane or cantilever, relying on the crane's movement or the cantilever's motion, which is not only cumbersome but also time-consuming for data acquisition. The fixed mode requires a combination of a gimbal and radar to scan the stockpile, with each measurement point measured individually. This results in inaccurate point cloud stitching and the need for interactive background element removal, making a single complete measurement time-consuming. Therefore, neither of these modes can achieve fully automated, accurate, and efficient measurement of large stockpiles. Summary of the Invention

[0004] To overcome the above-mentioned shortcomings, the present invention provides a fully automatic measurement method for material piles, so as to achieve fully automatic, accurate and efficient measurement of the volume and weight of material piles.

[0005] The technical solution adopted by this invention to solve its technical problem is as follows: A fully automated measurement method for material piles includes: Multiple measuring points are fixedly installed on the roof of the material yard, and each measuring point is equipped with a radar with a gimbal; the gimbal has a yaw joint and a pitch joint, and the yaw joint is close to the roof; the radar is fixedly installed below the gimbal; the radar and the gimbal are respectively connected to a data processing computer via Ethernet. The radar and pan-tilt unit simultaneously transmit instantaneous radar measurement data to the data processing computer via Ethernet. r , i ) and gimbal yaw and pitch data ( α , β );in r The distance from the radar center to the current surface point. i The scanning angle; α This refers to the yaw angle of the gimbal. β This refers to the instantaneous pitch angle of the gimbal; Determine the transformation process between the radar and gimbal's transmitted data and the coordinate system; Combined coordinate system transformation transforms individual radar measurements ( r , i Convert the points on the stockpile surface to homogeneous coordinates in the system coordinate system. P 0 = ( x , y , z ,1); For measurement points k ,because i Uniform sampling within a given interval results in a discrete planar curve for each radar measurement; the yaw angle of the gimbal is continuously varied by remotely transmitting angle positioning commands. α Or pitch angle β The scanning process yields a series of continuously changing curves, which are then converged to form a local point cloud. C k ={( x , y , z ,1)}, this measurement point k The corresponding local point cloud reflects the three-dimensional shape of a local area of ​​the stockpile; The ICP algorithm is used to register adjacent local point clouds, and then the adjacent point clouds are stitched together sequentially to generate the overall point cloud of the material pile. C ; Divide the ground area of ​​the material yard into a regular grid of N rows and M columns Ω = {( i , j )| i =0,…,M-1, j =0,…,N-1}, Pile cloud C The projection on the ground is C p ={( x ,0, z )|( x , y , z ,1)∈ C}, grid points ( i , j The nearest neighbor of a grid point is defined as the set of point clouds that fall within the grid points. Grid points ( i , j elevation h ( i , j ) is defined as the average of the absolute values ​​of the second coordinate components of the nearest neighbors; where D x and D z This indicates the spacing of the regular grid in the width and length directions of the material yard; Indicates rounding down; Based on the regular grid Ω and the elevation of the grid points, the elevation model DEM of the stockpile is defined as a set of triples: DEM = {( i , j , h ( i , j ))}|( i , j )∈Ω}; After filling holes and smoothing edges, an elevation model (DEM) is used to calculate the volume V and weight of the stockpile. .

[0006] Optionally, the stockpile includes unpackaged block, granular, and powdered materials piled together.

[0007] Optionally, the material yard is rectangular; the upper part of the rectangular material yard is supported by columns to form a roof; a track for a crane to travel on is set in the center of the material yard ground, and the left and right sides of the track are material stacking areas; multiple measuring points are deployed on the supports on both sides of the roof, and the union of the scanning areas of all measuring points completely covers the material yard ground.

[0008] Optionally, before data acquisition, the slewing mechanism of the trolley is set to return to the zero position, so that its cantilever is located within the track.

[0009] Optionally, the process of determining the transformation between the radar and gimbal's transmitted data and the coordinate system specifically includes: The system coordinate system O0-X0Y0Z0, the gimbal yaw coordinate system O1-X1Y1Z1, the gimbal pitch coordinate system O2-X2Y2Z2, and the radar coordinate system O3-X3Y3Z3 are established respectively. The transformation between the coordinate systems is represented by a combination of translation and rotation.

[0010] Optionally, the combined coordinate system transformation converts individual radar measurements ( r , i Convert the points on the stockpile surface to homogeneous coordinates in the system coordinate system. P 0 = ( x , y , z ,1), specifically including: Using formula A single measurement value from the radar ( r , i Convert the points on the stockpile surface to homogeneous coordinates in the system coordinate system. P 0 = ( x , y , z ,1); where d L This indicates the distance from the projection of the gimbal onto the ground to the left boundary of the material yard. d HThis indicates the vertical distance from the gimbal's yaw joint to the ground. d B This indicates the distance from the projection of the gimbal onto the ground to the rear boundary of the material yard; d py This indicates the vertical distance between the yaw and pitch joints of the gimbal; T ( d x , d y , d z ) is a translation matrix, and the parameter ( d x , d y , d z These represent the translation amounts of the three coordinate axes; R axis ( d () represents a rotation matrix about a coordinate axis. axis = x , y The parameter δ is the rotation angle.

[0011] Optionally, the ICP algorithm is used to register adjacent local point clouds, and the adjacent point clouds are stitched together sequentially to generate the overall point cloud of the material pile. C Specifically, it includes: Set up 4 measurement points k The point clouds corresponding to =1,2,3,4 are respectively C k , C 1 and C 2. There is an overlapping region in the Z0 direction. C 2 and C 3. There is an overlapping area in the X0 direction. C 3 and C 4. There is an overlapping region in the Z0 direction. Based on the overlapping region, the ICP algorithm is used to calculate the adjacent point clouds. C k+1 arrive C k Transformation matrix ; Adjacent point clouds are stitched together sequentially to generate the overall point cloud of the material pile. .

[0012] Optionally, the coordinate components of the stockpile point ( x , y , z )satisfy: ;in L and W For the length and width of the material yard, HThe height of the lower edge of the ceiling support from the ground. This refers to the track width.

[0013] Optionally, the patching and edge smoothing specifically include: Grid points are determined by observing whether the elevation value is zero or whether the nearest neighbor is an empty set. i , j Is it a cavity? Let the radius of the cavity be R. c If grid points ( i , j ( ) is a cavity, its elevation h ( i , j ) is approximately R c - The average elevation of all non-hole grid points within the neighborhood; If there are vertical steep slope-like jumps at some edges of the material pile, the elevation value of that part needs to be smoothed appropriately. First, treat the elevation model DEM as an image, use the edge detection operator to detect the outer edge of the material pile, and then use Gaussian filtering or mean filtering to smooth the area near the edge.

[0014] Optionally, the elevation model (DEM) is used to calculate the volume V and weight of the stockpile. Specifically, it includes: ; ; in m This refers to the density of the stockpile.

[0015] The beneficial effects of this invention are: This invention provides a fully automated measurement method for stockpiles, which deploys multiple measurement points (each measurement point includes a pan-tilt unit and a radar) on the roof of the stockyard. The radar and pan-tilt unit simultaneously transmit instantaneous radar measurement data to a data processing computer via Ethernet. r , i ) and gimbal yaw and pitch data ( α , β The volume V and weight of the stockpile can be obtained through fully automated calculations. This invention, once deployed, eliminates the need for on-site human intervention, automating the entire measurement process and saving labor costs. It also achieves automated data acquisition and point cloud processing, resulting in high-precision point cloud stitching for each measurement point, thus improving measurement accuracy and efficiency. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the measurement point deployment in the rectangular material yard described in an embodiment of the present invention; Figure 2This is a schematic diagram of the gimbal yaw coordinate system, gimbal pitch coordinate system, and radar coordinate system as described in an embodiment of the present invention; Figure 3 This is a schematic diagram of the material yard ground as described in an embodiment of the present invention; Figure 4 This is a schematic diagram of the cross-section of the material yard as described in an embodiment of the present invention; Referring to the accompanying drawings, the following explanations are provided: 1. Rectangular material yard; 2. Columns; 3. Canopy; 4. Track; 5. Material stacking area; 6. Measurement point; O0-X0Y0Z0, system coordinate system; O1-X1Y1Z1, gimbal yaw coordinate system; O2-X2Y2Z2, gimbal pitch coordinate system; O3-X3Y3Z3, radar coordinate system. Detailed Implementation

[0017] The following describes the specific implementation method in detail using a rectangular material yard with a roof as an example. The specific implementation method described herein is only for intuitive explanation of the present invention and is not intended to limit the present invention. The present invention can be implemented in other forms, such as a circular material yard with a roof.

[0018] See Figure 1 In this embodiment, the upper part of the rectangular material yard 1 is supported by columns 2 to form a roof 3. A track 4 for a traveling vehicle is laid in the center of the material yard ground, and the left and right sides of the track are material stacking areas 5. Multiple measuring points 6 are deployed on the supports on both sides of the roof, and each measuring point is equipped with a 2D line lidar (hereinafter referred to as lidar or laser radar) with a pan-tilt unit. The lidar moves with the pan-tilt unit. By automatically adjusting the yaw or pitch angle of the pan-tilt unit, the lidar can scan the local area below the installation position to obtain a local point cloud. The material pile of the present invention includes a large amount of unpackaged block, granular, and powdery materials piled together, such as coal or ore. The pan-tilt unit of the present invention has a yaw joint and a pitch joint. The pan-tilt unit is fixedly installed on the roof of the material yard, and the yaw joint is close to the roof. The lidar is fixedly installed below the pan-tilt unit. The lidar and the pan-tilt unit are connected to the data processing computer via Ethernet.

[0019] Before data acquisition, the slewing mechanism of the gantry crane is set back to the zero position, ensuring its cantilever is within the track and does not interfere with the radar's scanning of the material pile. The number of measurement points is related to the size of the rectangular material yard. Generally, a larger material yard requires more measurement points, while a smaller material yard requires fewer. A basic principle is that the union of the scanning areas of all measurement points must completely cover the ground of the material yard.

[0020] The radar and gimbal simultaneously transmit instantaneous radar measurement data and gimbal yaw and pitch data to the data processing computer via Ethernet. A single radar measurement value is ( r , i ),in r The distance from the radar center to the current surface point. i The scanning angle is [value]. The gimbal's yaw angle is [value]. α The instantaneous pitch angle of the gimbal is β .

[0021] First, determine the transformation process between the radar and gimbal's transmitted data and the coordinate system. For example... Figure 1 and Figure 2 As shown, the system coordinate system O0-X0Y0Z0, the gimbal yaw coordinate system O1-X1Y1Z1, the gimbal pitch coordinate system O2-X2Y2Z2, and the radar coordinate system O3-X3Y3Z3 are established respectively. The transformation between the coordinate systems can be represented by a combination of translation and rotation.

[0022] make T ( d x , d y , d z ) is a translation matrix, and the parameter ( d x , d y , d z ) represent the translation amounts of the three coordinate axes. R axis ( d () represents a rotation matrix about a coordinate axis. axis = x , y ,parameter d Let be the rotation angle, then we have: .

[0023] Taking a single measurement point as an example, such as Figure 2 As shown, let the single radar measurement value be ( r , i ), r The distance from the radar center to the current surface point. i The scanning angle. Homogeneous coordinates of the gimbal pitch coordinate system. P 2 can be expressed in homogeneous coordinates of the radar coordinate system as follows: ; in,( d x , d y , d z ) represents the three-dimensional coordinates (hereinafter referred to as coordinates) of the radar coordinate system origin in the gimbal elevation coordinate system, while ( d x ,d y , d z ,1) represents the four-dimensional homogeneous coordinates (hereinafter referred to as homogeneous coordinates) of the radar coordinate system origin in the gimbal elevation coordinate system. These three parameters ( d x , d y , d z The symbol () represents the radar's relative position to the gimbal, which can be determined during installation. The radar must be installed symmetrically with respect to the gimbal. d x =0.

[0024] Homogeneous coordinates of the gimbal yaw coordinate system P 1. Homogeneous coordinates of the gimbal pitch coordinate system P 2 means: ; in, d py This indicates the vertical distance between the yaw and pitch joints of the gimbal; it is a parameter of the gimbal itself and can be measured before installation. (0, d py ,0) represents the coordinates of the origin of the gimbal's pitch coordinate system in the yaw coordinate system. β This refers to the instantaneous pitch angle of the gimbal.

[0025] Homogeneous coordinates in the system coordinate system P 0 can be derived from the homogeneous coordinates in the gimbal yaw coordinate system. P 1 is represented as: ; in,( d L ,- d H , d B () represents the installation location of the measurement point, indicating the coordinates of the origin of the gimbal yaw coordinate system in the system coordinate system. α This refers to the yaw angle of the gimbal. For example... Figure 1 and Figure 3 As shown, d L This indicates the distance from the projection of gimbal 1 onto the ground to the left boundary of the material yard. d B This represents the distance from the projection of gimbal 1 onto the ground to the rear boundary of the material yard. d H These three values ​​represent the vertical distance from the gimbal's yaw joint to the ground; they are determined during equipment installation.

[0026] Combining the aforementioned coordinate system transformations, a single measurement value from the radar ( r , i This can be converted into the homogeneous coordinates of points on the stockpile surface in the system coordinate system. P 0 = ( x , y , z ,1), that is: .

[0027] For measurement points k ,because i Uniform sampling within a given interval results in a discrete planar curve for each radar measurement. The yaw angle of the gimbal is continuously varied by remotely transmitting angle positioning commands. α Or pitch angle β It can scan to obtain a series of continuously changing curves, and converge these curves to form a local point cloud. C k ={( x , y , z ,1)}, this measurement point k The corresponding local point cloud reflects the three-dimensional shape of a local area of ​​the stockpile.

[0028] The radar's high location and the complex structure of the material yard walls made manual ranging difficult on-site. Errors in the installation position parameters led to significant errors in the conversion of radar measurements to system coordinates. Therefore, directly merging local point clouds to generate the overall point cloud was problematic. It cannot effectively represent the geometry of the material pile. This implementation method uses the ICP algorithm to register adjacent local point clouds, achieving accurate stitching of local point clouds.

[0029] like Figure 3 As shown, taking four measurement points as an example, Figure 3 In the diagram, 1, 2, 3, and 4 represent the projections of gimbal 1, gimbal 2, gimbal 3, and gimbal 4, respectively, and the point clouds corresponding to the measurement points are respectively... C k , k =1,2,3,4, and more measurement points can be handled similarly. C 1 and C 2. There is an overlapping region in the Z0 direction. C 2 and C 3. There is an overlapping area in the X0 direction. C 3 and C 4. There is an overlapping region in the Z0 direction. Based on the overlapping region, the ICP algorithm can be used to calculate the adjacent point clouds. C k+1 arrive C k Transformation matrix By sequentially stitching together adjacent point clouds, the overall point cloud of the material pile can be generated according to the following formula. C : .

[0030] The scanning process may capture background elements such as walls, ceilings, cranes, and tracks of the material yard into a point cloud. This implementation method removes these background elements based on the geometric model of the material yard. Figure 3 and Figure 4 As shown, L and W For the length and width of the material yard, H The height of the lower edge of the ceiling support from the ground. W 0 represents the track width. Therefore, the coordinate components of the stockpile point ( x , y , z The following conditions must be met: ; The constant in the above formula H It can be promoted as about x Functions: H = H ( x ), x ∈[0, W It supports more flexible rejection functionality.

[0031] The vertical span of a rectangular stockpile is much smaller than its horizontal span, and the geometric surface of the stockpile can be represented using an elevation model. The ground... W × L The rectangular area is divided into a regular grid of N rows × M columns Ω={( i , j )| i =0,…,M-1, j =0,…,N-1} D x and D z This indicates the spacing of the regular grid along the width and length of the stockpile. Stockpile point cloud. C The projection on the ground is C p ={( x ,0, z )|( x , y , z ,1)∈ C}.like , If the value is rounded down, it is called a point ( x , y , z ) or its projection point (x ,0, z ) Falling into grid point ( i , j ). Grid points ( i , j The nearest neighbor of a grid point is defined as the set of point clouds that fall within the grid points. Grid points ( i , j elevation h ( i , j The elevation of a grid point is defined as the average of the absolute values ​​of the second coordinate components of its nearest neighbors. If the nearest neighbors of a grid point are an empty set, the elevation of that grid point is assigned a value of zero. Based on the regular grid Ω and the elevations of the grid points, the elevation model DEM of the stockpile can be formally defined as a set of triples: DEM = {( i , j , h ( i , j ))}|( i , j )∈Ω}.

[0032] The non-uniformity of the gimbal's mechanical movement leads to uneven point cloud distribution. A small number of voids may exist within the material pile. Whether a grid point is a void can be determined by observing whether its elevation value is zero or whether its nearest neighbor is an empty set. Let the radius of the void be R. c If grid points ( i , j () is a cavity, its elevation h ( i , j It can be approximated as R c - The average elevation of all non-hole grid points within the neighborhood. Parameter R c The value is related to the distribution of hole size. After collecting point clouds on site, it can be set as appropriate, and 0.1 meters is recommended.

[0033] Occlusion can cause gaps at some edges of the material pile. If there is a visually noticeable vertical abrupt change at this edge, the elevation value of that area needs to be smoothed appropriately. First, treat the elevation model (DEM) as an image and use an edge detection operator (such as the Canny operator) to detect the outer edges of the material pile. Then, use Gaussian filtering or mean filtering to smooth the area near the edges. A 7×7 filter window size is recommended.

[0034] After filling holes and smoothing edges, the elevation model (DEM) can be used to calculate the volume (V) and weight of the stockpile. ,set up m Let the density of the stockpile be: ; .

[0035] Furthermore, each sub-rectangular grid of the elevation model DEM {( i , j ), ( i +1, j ), ( i +1, j +1), ( i , j +1)} corresponds to the four vertices: ( i · D x ,- h ( i , j ), j · D z ), (( i +1)· D x ,- h ( i +1, j ), j · D z ), (( i +1)· D x ,- h ( i +1, j +1),( j +1)· D z )and( i · D x ,- h ( i , j +1),( j +1)· D z Based on the diagonal, the four vertices can be divided into two triangles. Traversing the mesh Ω generates a triangular facet model of the material pile, realizing the 3D modeling of the material pile.

[0036] This invention provides a fully automated measurement method for material piles. Multiple measurement points (each measurement point includes a pan-tilt unit and a 2D lidar) are deployed on the roof of the material yard. The data acquisition process for each measurement point is concurrently controlled by software, and the transformation process between the radar and pan-tilt unit's transmitted data and the coordinate system is determined. The local point clouds of adjacent measurement points are then stitched together to obtain a point cloud model of the material pile. CThe process involves using the geometric model of the material yard to remove background elements, projecting the point cloud model onto the ground to obtain the elevation model (DEM) of the material pile, filling in voids in the pile and smoothing abrupt edges, and finally using the elevation model to calculate the volume (V) and weight of the pile. This invention enables 3D modeling. Once deployed, the entire measurement process is automated, eliminating the need for on-site human intervention and saving labor costs. It achieves automated data acquisition and point cloud processing, with high accuracy in stitching the point clouds from each measurement point, significantly improving measurement accuracy and efficiency.

[0037] Many specific details have been set forth in the foregoing description to provide a thorough understanding of the present invention. However, the above description is merely a preferred embodiment of the present invention, and the present invention can be implemented in many other ways different from those described herein. Therefore, the present invention is not limited to the specific embodiments disclosed above. Furthermore, any person skilled in the art can make many possible variations and modifications to the technical solutions of the present invention, or modify them into equivalent embodiments, using the methods and techniques disclosed above, without departing from the scope of the present invention. Any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention, without departing from the content of the present invention, shall still fall within the protection scope of the present invention.

Claims

1. A method for full automatic measurement of a stockpile, characterized in that, The application relates to a method for measuring the height of a stockpile. A plurality of measuring points are fixedly arranged on the top of the stockpile, and each measuring point is provided with a radar with a gimbal; The gimbal has a yaw joint and a pitch joint, and the yaw joint is close to the top; the radar is fixedly arranged below the gimbal; the radar and the gimbal are respectively connected to a data processing computer through an Ethernet. The radar and the gimbal simultaneously send instant radar measurement data ( Determine the conversion process of the backhaul data of the radar and the gimbal and the coordinate system. , The stockpile comprises unpackaged blocky, granular and powdery materials piled together. ) and yaw and pitch data ( α , β ) of the gimbal to the data processing computer through Ethernet; wherein The stockpile is rectangular; the upper part of the rectangular stockpile is supported by a stand column; a track for a travelling car is arranged in the center of the ground of the stockpile, and the left and right sides of the track are the stockpile areas; a plurality of measuring points are arranged on the supports on the two sides of the top, and the union of the scanning areas of all the measuring points completely covers the ground of the stockpile. is the distance from the radar center to the current surface point, Before data collection, the rotating mechanism of the travelling car is set to return to the zero position, so that the cantilever of the travelling car is located in the track. is the scanning angle; α is the yaw angle of the gimbal, β is the instant pitch angle of the gimbal;​​ The conversion process of the backhaul data of the radar and the gimbal and the coordinate system comprises the following steps. Combined coordinate system transformation transforms individual radar measurements ( System coordinate system O0-X0Y0Z0, gimbal yaw coordinate system O1-X1Y1Z1, gimbal pitch coordinate system O2-X2Y2Z2 and radar coordinate system O3-X3Y3Z3 are respectively established, and the transformation between the coordinate systems is represented by the combination of translation and rotation. , The hole filling and edge smoothing specifically comprises the following steps. Convert the points on the stockpile surface to homogeneous coordinates in the system coordinate system. P 0 = ( x , y , z ,1); For the measurement point k , since If there is a vertical abrupt slope phenomenon at the edge of the part of the stockpile, the elevation value of the part needs to be appropriately smoothed; first, the elevation model DEM is regarded as an image, the peripheral edge of the stockpile is detected by using an edge detection operator, and then the region near the edge is smoothed by using Gaussian filtering or mean filtering. , the result of each measurement of the radar is a discrete planar curve; by continuously varying the yaw angle of the pan-tilt head through remote transmission of angle positioning instructions α or the pitch angle β , a series of continuously varying curves are scanned, and these curves are aggregated to form a local point cloud C k = {( x , y , z , 1)}, the corresponding local point cloud of the measurement point k reflects the three-dimensional shape of the local area of the material pile; ICP algorithm is adopted to register adjacent local point clouds, and the adjacent point clouds are sequentially spliced to generate the overall point cloud of the material pile C ; Divide the ground area of ​​the material yard into a regular grid of N rows and M columns Ω = {( i , j )| i =0,…,M-1, j =0,…,N-1}, Pile cloud C The projection on the ground is C p ={( x ,0, z )|( x , y , z ,1)∈ C }, grid points ( i , j The nearest neighbor of a grid point is defined as the set of point clouds that fall within the grid points. Grid points ( i , j elevation h ( i , j ) is defined as the average of the absolute values ​​of the second coordinate components of the nearest neighbors; where D x and D z This indicates the spacing of the regular grid in the width and length directions of the material yard; Indicates rounding down; Based on the regular grid Ω and the elevation of the grid points, the elevation model DEM of the stockpile is defined as a set of triples: DEM = {( i , j , h ( i , j ))}|( i , j )∈Ω}; After the holes and the edges are smoothed, the volume V and the weight of the pile are calculated using the elevation model DEM .

2. The method of claim 1, wherein The application relates to a method for measuring the height of a stockpile.

3. The method of claim 1, wherein ​ 4. The method of claim 3, wherein ​ 5. The method of claim 1, wherein ​ ​ 6. The method of claim 5, wherein The combined coordinate system transformation converts individual radar measurements ( ​ , ​ Convert the points on the stockpile surface to homogeneous coordinates in the system coordinate system. P 0 = ( x , y , z ,1), specifically including: Using formula A single measurement value from the radar ( ​ , ​ Convert the points on the stockpile surface to homogeneous coordinates in the system coordinate system. P 0 = ( x , y , z ,1); where d L This indicates the distance from the projection of the gimbal onto the ground to the left boundary of the material yard. d H This indicates the vertical distance from the gimbal's yaw joint to the ground. d B This indicates the distance from the projection of the gimbal onto the ground to the rear boundary of the material yard; d py This indicates the vertical distance between the yaw and pitch joints of the gimbal; T ( d x , d y , d z ) is a translation matrix, and the parameter ( d x , d y , d z These represent the translation amounts of the three coordinate axes; R axis ( ​ () represents a rotation matrix about a coordinate axis. ​ = x , y The parameter δ is the rotation angle.

7. The method of claim 6, wherein The ICP algorithm is used to register adjacent local point clouds, and the adjacent point clouds are spliced in sequence to generate the overall point cloud of the material pile C , and specifically comprises: Set 4 measurement points k = 1, 2, 3, 4, the corresponding point clouds are respectively C k , C 1 and C 2 have an overlapping area in the Z0 direction, C 2 and C 3 have an overlapping area in the X0 direction, C 3 and C 4 have an overlapping area in the Z0 direction, based on the overlapping area, the ICP algorithm is used to calculate the adjacent point clouds C k+1 to C k the transformation matrix ; The adjacent point clouds are spliced in sequence to generate the overall point cloud of the material pile .

8. The method of claim 7, wherein The coordinate components of the stockpile point ( x , y , z )satisfy: ;in L and W For the length and width of the material yard, H The height of the lower edge of the ceiling support from the ground. This refers to the track width.

9. The method of claim 7, wherein the step of measuring the height of the material pile is performed by a laser sensor. ​ Grid points are determined by observing whether the elevation value is zero or whether the nearest neighbor is an empty set. i , j Is it a cavity? Let the radius of the cavity be R. c If grid points ( i , j ( ) is a cavity, its elevation h ( i , j ) is approximately R c - The average elevation of all non-hole grid points within the neighborhood; ​ 10. The method of claim 9, wherein The use of an elevation model DEM to calculate the volume V and weight of the pile , specifically comprising: ； ； wherein ​ is the density of the pile.

Images