Method for detecting shape of material on belt conveyor by adopting two-dimensional laser scanner

Abstract

The invention discloses a method for detecting the shape of a material on a belt conveyor through a two-dimensional laser scanner. The method comprises the following steps that S1, the two-dimensional laser scanner is installed; s2, calibrating a carrier roller model of the sealing-tape machine; s3, the material shape and the mass center of the sealing-tape machine are detected; and S4, the blanking position of a material pouring plate of the sealing-tape machine is adjusted, and the material loss is calculated. The material falling position of the material pouring plate is accurately controlled, the material loss of the belt conveyor in the conveying process is detected, and meanwhile the defect that sharp foreign matter is mixed in materials can be detected.

Description

Technical Field

[0001] This invention relates to iron and steel metallurgy technology, and more specifically, to a method for detecting the shape of materials on a belt conveyor using a two-dimensional laser scanner. Background Technology

[0002] Belt conveyors are the primary material handling equipment in steel plants. Material transported by the upstream belt conveyor falls onto the downstream belt conveyor's belt via guide plates in the tail chute. If the material drop point deviates significantly from the belt conveyor's center, the belt is prone to misalignment and material spillage, severely impacting normal material transport. Furthermore, material loss occurs during belt conveying due to material drop or spillage; this loss must be controlled within a certain range.

[0003] In existing patent applications, such as patent publication number CN109665284A, a belt conveyor load monitoring instrument and system are disclosed. The belt conveyor load monitoring instrument is composed of a laser ranging module, a laser control processing module, and a rotating mechanism. The monitoring instrument scans the raw coal on the belt conveyor to obtain the load distribution per unit length and the current operating speed of the belt conveyor, and finally adjusts the operating speed of the belt conveyor.

[0004] For example, patent publication number CN 213455719U discloses a dynamic metering device for bulk materials on a belt conveyor. By setting a laser scanner and a camera device above the belt conveyor, the laser scanner generates a linear light source that illuminates the surface of the material above the conveyor belt to form a laser stripe outline. Then, the camera takes a real-time picture of the laser stripe outline to obtain the material outline height. The controller uses linear interpolation to calculate the material cross-sectional area and flow rate based on the data collected by the camera.

[0005] Neither of the two patented technologies mentioned above involves the use of a two-dimensional laser scanner to detect the centroid of the material conveyed by the belt conveyor, to guide the control of the unloading position of the unloading plate by the centroid change trend, or to detect the loss of material conveyed by the belt conveyor and the detection of sharp foreign objects mixed in the material. Summary of the Invention

[0006] In view of the shortcomings of the existing technology, the purpose of this invention is to provide a method for detecting the shape of materials on a belt conveyor using a two-dimensional laser scanner, so as to achieve precise control of the material dropping position of the discharge plate, detect material loss during the conveying process of the belt conveyor, and at the same time, detect defects such as sharp foreign objects mixed in the material.

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] A method for detecting the shape of materials on a conveyor belt using a two-dimensional laser scanner includes the following steps:

[0009] S1, Install a 2D laser scanner;

[0010] S2, model of idler roller for calibrating conveyor belt machine;

[0011] S3, Detect the material shape of the conveyor belt and calculate the centroid;

[0012] S4, Adjust the material discharge position of the conveyor belt's discharge plate and calculate material loss.

[0013] Preferably, step S1 is as follows:

[0014] The two-dimensional laser scanner is positioned directly above the idler roller of the conveyor belt machine;

[0015] The scanning line plane of the two-dimensional laser scanner is perpendicular to the center line of the belt of the conveyor belt machine and intersects the upper contour line of the idler roller below the belt;

[0016] When there is no material on the belt, the 0-degree scanning line of the two-dimensional laser scanner is perpendicular to and intersects the midpoint of the idler roller.

[0017] Preferably, step S2 is as follows:

[0018] A Cartesian coordinate system is established with the center of the scan line plane of the two-dimensional laser scanner as the 0 point;

[0019] In the Cartesian coordinate system, the positive direction of the y-axis is vertically downward, and the positive direction of the x-axis is perpendicular to the center line of the belt and away from the center line.

[0020] When there is no material on the belt and it is stationary, the two-dimensional laser scanner performs a line scan laser scan to acquire idler roller model data.

[0021] Preferably, step S3 is as follows:

[0022] During material conveying, the two-dimensional laser scanner performs one scan according to the sampling cycle to obtain point cloud data for a single scan. The point cloud data is then filtered to obtain material contour point cloud data. Combined with the idler roller model, a closed cross-sectional shape of the material scanned by the single point cloud data is constructed. The area of ​​the closed cross-sectional shape of the material scanned by the single point cloud data, the centroid, and the offset between the centroid and the centerline of the belt are calculated.

[0023] Preferably, the point cloud data is filtered as follows:

[0024] Each single scan yields a set of data points a in polar coordinates along the upper outer contour of the idler roller. The set of data points a is then converted into a set of data points b in the Cartesian coordinate system. After performing discrete point filtering and conditional filtering on the set of data points b, a set of data points c is obtained where all points have an angle greater than the angle between the hypotenuse vertex of one side of the idler roller and the x-axis, and less than the angle between the hypotenuse vertex of the other side of the idler roller and the positive x-axis.

[0025] When there is no material on the belt, the idler model is extracted from the data point set c and used as the reference model for subsequent detection.

[0026] Preferably, the methods for extracting the idler model based on the idler angle and the processing speed of the point cloud data specifically include linear fitting of the idler model, polynomial fitting of the idler model, and direct use of the data point set.

[0027] Preferably, the linear fitting idler roller model specifically includes:

[0028] For the data point set c, a straight line fitting method is used to obtain the fitted straight line f1 of the hypotenuse of the other side of the idler roller, the fitted straight line f2 of the hypotenuse of one side of the idler roller, and the fitted straight line f3 of the bottom of the idler roller. The coordinates P1′(x) of the intersection point between the hypotenuse of one side of the idler roller and the bottom of the idler roller are then calculated. c1 ′,y c1 The coordinates of the intersection point P1(x′) between the hypotenuse of the other side of the idler roller and the bottom of the idler roller are given. c1 ,y c1 The fitted line f1, the fitted line f2 and the fitted line f3, and the intersection point P1(x) c1 ,y c1 The intersection point P1′(x) c1 ′,y c1 The straight-line fitting model d is formed by the ′) of the idler roller.

[0029] In the material conveying state, the two-dimensional laser scanner performs a single scan to obtain point cloud data p1{(x0,y0),(x1,y1),(x2,y2),...,(x n ,y n )}gather.

[0030] Preferably, the polynomial fitting idler roller model specifically includes:

[0031] For the data point set c, the function f(x) = a n x n +a n-1 x n-1 By fitting +......+a0, we obtain the idler roller model e;

[0032] For the idler roller model, a quadratic curve is used.

[0033] Preferably, the direct use of the data point set specifically includes:

[0034] The data point set c is directly used as the idler roller model g.

[0035] Preferably, the material shape contour extraction based on the linear fitting idler roller model specifically includes:

[0036] Remove all points in the data point set c whose distance to the fitted lines f1 and f2 is less than a set threshold h, thereby obtaining the point cloud dataset p2{(x0′,y0′),(x1′,y1′),(x2′,y2′),...,(x m ′,y m ′)};

[0037] The point cloud dataset p2 is combined with the intersection point coordinates P1′(x) c1 ′,y c1 The intersection point coordinates P1(x′) c1 ,y c1 Construct a new point cloud dataset p3{x0′,y0′),(x1′,y1′),(x2′,y2′),...,(x m ′,y m ′), (x c1 y c1 ),(x c1 ′,y c1 Starting from p3{x0′,y0′), following the order of data in the point cloud set, the data points in p3 are sequentially connected using fold lines in a clockwise direction to obtain the region of the closed graphic of the single point cloud data material scanning section.

[0038] Preferably, the material shape contour extraction based on the polynomial fitting of the idler roller mold specifically includes:

[0039] Remove all points from the data set c to fit the polynomial f(x) = a n x n +a n-1 x n-1 Points whose distance from a0 is less than a set threshold h are used to obtain the point cloud dataset p2{(x0′,y0′),(x1′,y1′),(x2′,y2′),...,(x m ′,y m Then extract f(x) = a')}. n xn +a n-1 x n-1 +......+a0, the fitted curve lies between x0′ and x m Point cloud dataset p4{(x) between ' p0 ′,y p0 ′),(x p1 ′,y p1 ′),(x p2 ′,y p2 ′),...,(x pm ′,y pm The point cloud datasets p2 and p4 are combined to form a new point cloud dataset p3{(x0′,y0′),(x1′,y1′),(x2′,y2′),...,(x′)}, where p2 and p4 are combined to form a new point cloud dataset p3{(x0′,y0′),(x1′,y1′),(x2′,y2′),...,(x′)}. m ′,y m ′),(x pm ′,y pm ′),...,(x p0 ′,y p0 ′)}, thereby obtaining the region of the closed graphic of the material scan section of the single point cloud data.

[0040] Preferably, the material shape contour extraction based on the direct use of the data point set specifically includes:

[0041] All point clouds within the overlap threshold h between the data point set c and the idler roller model g are removed to obtain the point cloud dataset p2{(x0′,y0′),(x1′,y1′),(x2′,y2′),...,(x m ′,y m Then extract the idler roller model g between x0′ and x′. m Point cloud dataset p4{(x) between ' p0 ′,y p0 ′),(x p1 ′,y p1 ′),(x p2 ′,y p2 ′),...,(x pm ′,y pm The point cloud datasets p2 and p4 are combined to form a new point cloud dataset p3{(x0′,y0′),(x1′,y1′),(x2′,y2′),...,(x′)}, where p2 and p4 are combined to form a new point cloud dataset p3{(x0′,y0′),(x1′,y1′),(x2′,y2′),...,(x′)}. m ′,y m ′),(x pm ′,y pm ′),...,(x p0 ′,y p0′)}, thereby obtaining the region of the closed graphic of the material scan section of the single point cloud data.

[0042] Preferably, the calculation of the region and area of ​​the closed shape of the material scan section of the single point cloud data is as follows:

[0043] The point cloud dataset p3 is equivalently represented according to the original point cloud order as p3{p0,p1,p2,...,p...} n};

[0044] With p0 as the origin, p i and p i+1 For the other two vertices of the triangle, according to p0p i p i+1 p0 sequentially constructs vector triangles, dividing the region of the closed graphic of the single point cloud data material scan section into n-2 vector triangles;

[0045] Calculate p0p using the cross product of vectors. i p i+1 The area of ​​triangle p0 is s i The sum of the areas of all the triangles is the area s of the closed shape of the cross-section of the material scanned in a single point cloud data session, which is s = ∑s i .

[0046] The centroid cx of each triangle i =(p0.x+p i .x+p i+1 .x) / 3,cy i =(p0.y+p i .y+p i+1 .y) / 3, the centroid of the closed region 18 is xm=(∑cx i *s i ) / ∑s i ,y m =(∑cy i *s i ) / ∑s i .

[0047] Preferably, adjusting the material discharge position of the tape conveyor's feeding plate in step S4 specifically includes:

[0048] Calculate the position of the centroid m(x) m ,y m The deviation d between the centerline position of the conveyor belt and the centerline position of the conveyor belt machine m ;

[0049] The maximum deviation between the centroid of the closed section of the material scanned from the single point cloud data and the centerline of the conveyor belt is d. max ;

[0050] The deviation adjustment threshold value between the centroid position of the closed shape of the material scan section of the single point cloud data and the centerline position of the conveyor belt is d. Δ1 ;

[0051] If d m ≤d Δ1 This indicates that the material drop point of the pouring plate is within the normal range and no adjustment is required;

[0052] If d m ≥d max This indicates that the material drop point of the discharge plate is too far off, and the machine should be stopped for inspection.

[0053] If dΔ1<d m <d max This indicates that the material drop point of the discharge plate is abnormal.

[0054] Preferably, the calculation of material loss in step S4 specifically includes:

[0055] The scanning time of the two-dimensional laser scanner is t. Within the time length T, calculate the conveying speed v of the tape machine and the volume conveyed on the tape machine:

[0056]

[0057] A two-dimensional laser scanner is installed at the head, middle, and tail of the conveyor belt. By calculating the difference in the conveying volume V of the conveyor belt within a time length T, and combining it with the bulk density ρ of the material, the material drop and spillage loss during the conveying process is evaluated. The ratio of the material loss during the conveying process to the rated conveying capacity of time T is calculated to evaluate the material conveying loss rate.

[0058] The present invention provides a method for detecting the shape of materials on a belt conveyor using a two-dimensional laser scanner. This method is based on a two-dimensional laser scanner to detect the shape of materials conveyed by a belt conveyor in a steel plant. It can then detect the cross-sectional area of ​​a single scan of the material above a designated idler roller, the position of the centroid, and the time trend of the centroid position change, thereby controlling the material drop position of the unloading plate. At the same time, by setting two-dimensional laser scanners at the head and tail of the conveyor and combining them with the material conveying speed, it is possible to detect material loss during the belt conveyor process. Furthermore, by detecting abrupt changes in the curvature of the material shape, it can be used to detect defects such as sharp foreign objects mixed in with the material. Attached Figure Description

[0059] Figure 1 This is a schematic diagram of the installation and calibration of the two-dimensional laser scanner in the detection method of the present invention;

[0060] Figure 2This is a schematic diagram of material shape detection using a two-dimensional laser scanner in the detection method of this invention;

[0061] Figure 3 This is a schematic diagram of the centroid detection of the material shape in the detection method of the present invention;

[0062] Figure 4 This is a schematic diagram of an embodiment of the detection method of the present invention. Detailed Implementation

[0063] To better understand the above-mentioned technical solutions of the present invention, the technical solutions of the present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0064] Combination Figure 1 , Figure 2 As shown, the present invention provides a method for detecting the shape of materials on a conveyor belt using a two-dimensional laser scanner, comprising the following steps:

[0065] S1, Install 2D laser scanner 1;

[0066] The 2D laser scanner 1 is installed directly above the belt 7 of the conveyor belt machine. When there is no material present, the scanning line plane 5 of the 2D laser scanner 1 (e.g., ...) Figure 1 The shaded area shown is perpendicular to the conveyor center line 3 of the belt conveyor and intersects with the outer contour line 10 of the trapezoidal idler roller 4. The installation height of the 2D laser scanner 1 needs to ensure that the scanning range of its scanning line plane 5 covers the top boundary point 11 of the upper inclined side 23 and the top boundary point 13 of the upper inclined side 22 of the trapezoidal idler roller 4. The zero-degree scanning line 9 of the 2D laser scanner 1 intersects with the trapezoidal idler roller 4 at the center point 8 of the bottom idler roller, which is on the conveyor center line 3 of the belt conveyor.

[0067] S2, model of idler roller for calibrating conveyor belt machine;

[0068] A Cartesian coordinate system is established with the center 2 of the scanning plane 5 of the two-dimensional laser scanner as the 0 point. The y-axis of the Cartesian coordinate system coincides with the zero-degree scanning line 9 of the two-dimensional laser scanner 1 and is vertically downward as the positive direction. The x-axis of the Cartesian coordinate system is perpendicular to the conveyor center line 3 of the conveyor belt and is far away from the conveyor center line 3 of the conveyor belt as the positive direction.

[0069] When there is no material on the belt 7 and it is stationary, the two-dimensional laser scanner 1 performs a line scan laser scan to acquire the idler model data.

[0070] S3, detects the shape and centroid of the material on the conveyor belt;

[0071] In the material conveying state, the two-dimensional laser scanner 1 performs one scan according to the sampling cycle to obtain the point cloud data of a single scan. The point cloud data is then filtered to obtain the material contour point cloud data. Combined with the idler roller model, a closed shape of the material scan section of the single point cloud data is constructed. The area of ​​the closed shape of the material scan section of the single point cloud data, the centroid, and the offset between the centroid and the center line of the belt are calculated.

[0072] The specific steps for filtering point cloud data are as follows:

[0073] Each single scan yields a set of data points a in polar coordinates along the outer contour 10 of the upper idler roller 4 in a material-free state or the material contour 17 in a material-containing state. The data point set a is then converted to a set of data points b in Cartesian coordinates. After performing discrete point filtering and conditional filtering on the data point set b, a set of data points c is obtained where the angle between all points and the positive x-axis is greater than the angle 20 between the vertices 11 of the hypotenuse 23 of the trapezoidal idler roller 4 and the x-axis, and less than the angle 21 between the vertices 13 of the hypotenuse 22 of the trapezoidal idler roller 4 and the positive x-axis.

[0074] When there is no material on belt 7, the idler model is extracted through data point set c and used as the benchmark model for subsequent detection.

[0075] Based on the roller angle and the point cloud data processing speed, the methods for extracting roller models include linear fitting of the roller model, polynomial fitting of the roller model, and direct use of the data point set. Details are as follows:

[0076] 1) The linear fitting idler roller model specifically includes:

[0077] For the data point set c, a straight line fitting method is used to obtain the fitted straight line f1 for the hypotenuse 22 of the trapezoidal idler, the fitted straight line f2 for the hypotenuse 23 of the trapezoidal idler, and the fitted straight line f3 for the bottom 24 of the idler. The coordinates P1′(x) of the intersection point 14 between the hypotenuse 22 and the bottom 24 are then calculated. c1 ′,y c1 The coordinates of the intersection point 12 between the hypotenuse 23 and the bottom 24 are P1(x). c1 ,y c1 ), fitted lines f1, f2 and f3, intersection point P1(x) c1 ,y c1 ), intersection point P1′(x) c1 ′,y c1 The straight-line fitting model d is formed by ′).

[0078] like Figure 2 As shown, in the material conveying state, the two-dimensional laser scanner 1 performs a single scan, obtaining point cloud data p1{(x0,y0),(x1,y1),(x2,y2),...,(xn ,y n )}gather.

[0079] 2) The polynomial fitting idler roller model specifically includes:

[0080] For a data point set c, the function f(x) = a n x n +a n-1 x n-1 By fitting +......+a0, we obtain the idler roller model e.

[0081] For the idler roller model, a quadratic curve is used.

[0082] 3) Directly using data point sets specifically includes:

[0083] The data point set c is directly used as the idler roller model g.

[0084] After extracting the idler roller model, material shape detection can be performed under material conveying conditions. For the data point set c obtained under each material conveying condition, there are three methods for extracting the material shape contour of the three idler roller models:

[0085] Method 1, material shape contour extraction based on linear fitting of the idler roller model, specifically includes:

[0086] Points in data set c whose distances to the fitted lines f1 and f2 are less than a set threshold h are removed, thus obtaining the point cloud dataset p2{(x0′,y0′),(x1′,y1′),(x2′,y2′),...,(x m ′,y m ′)}。 Where (x0′,y0′) corresponds to Figure 2 Position 15 of the standard number, (x m ′,y m ')correspond Figure 2 Position number 16.

[0087] Point cloud dataset p2 combined with the coordinates of intersection point 14 P1′(x) c1 ′,y c1 ′), the coordinates of intersection point 12 P1(x) c1 ,y c1 Construct a new point cloud dataset p3{x0′,y0′),(x1′,y1′),(x2′,y2′),...,(x m ′,y m ′), (x c1 y c1 ),(x c1 ′,y c1Starting from p3{x0′,y0′), following the order of data in the point cloud set, connect the data points in p3 sequentially in a clockwise direction using folding, thus obtaining the region 18 of the closed graphic of the material scan section of a single point cloud data (e.g., ′)}. Figure 2 (The shaded area shown).

[0088] Method 2, material shape contour extraction based on polynomial fitting of idler roller model, specifically includes:

[0089] Remove all data points from the set c to fit the polynomial f(x) = a n x n +a n-1 x n-1 Points whose distance to a0 is less than a set threshold h are used to obtain the point cloud dataset p2{(x0′,y0′),(x1′,y1′),(x2′,y2′),...,(x m ′,y m Then extract f(x) = a')}. n x n +a n-1 x n-1 +......+a0, the fitted curve lies between x0′ and x m Point cloud dataset p4{(x) between ' p0 ′,y p0 ′),(x p1 ′,y p1 ′),(x p2 ′,y p2 ′),...,(x pm ′,y pm The point cloud datasets p2 and p4 are combined to form a new point cloud dataset p3{(x0′,y0′),(x1′,y1′),(x2′,y2′),...,(x′)}, which is a combination of point cloud datasets p2 and p4. m ′,y m ′),(x pm ′,y pm ′),...,(x p0 ′,y p0 ′)}, thus obtaining the region 18 of the closed graphic of the material scan section of a single point cloud data (e.g. Figure 2 (The shaded area shown).

[0090] Method 3, material shape contour extraction based on directly using data point sets, specifically includes:

[0091] Remove all point clouds within the threshold h of the overlapping region between all data point sets c and the idler roller model g to obtain the point cloud dataset p2{(x0′,y0′),(x1′,y1′),(x2′,y2′),...,(x m ′,y m Then extract the idler roller model g between x0′ and x′. m Point cloud dataset p4{(x) between ' p0 ′,y p0 ′),(x p1 ′,y p1 ′),(x p2 ′,y p2 ′),...,(x pm ′,y pm Point cloud datasets p2 and p4 are combined to form a new point cloud dataset p3{(x0′,y0′),(x1′,y1′),(x2′,y2′),...,(x′)}, where p2 and p4 are combined to form a new point cloud dataset p3{(x0′,y0′),(x1′,y1′),(x2′,y2′),...,(x′)}. m ′,y m ′),(x pm ′,y pm ′),...,(x p0 ′,y p0 ′)}, thus obtaining the region 18 of the closed graphic of the material scan section of a single point cloud data (e.g. Figure 2 (The shaded area shown).

[0092] After completing the above data processing, the centroid and area of ​​the closed region 18 of the material cross-section are calculated for a single scan:

[0093] The point cloud dataset p3 is equivalently represented according to the original point cloud order as p3{p0,p1,p2,...,p...} n}, that is, p0 corresponds to (x0′, y0′), p1 corresponds to (x1′, y1′), and so on, p n Corresponding to (x) p0 ′,y p0 ′).

[0094] With p0 as the origin, p i and p i+1 For the other two vertices of the triangle, according to p0p i p i+1 p0 sequentially constructs vector triangles, dividing the closed region of the material scan section of a single point cloud data into n-2 vector triangles.

[0095] Calculate p0p using the cross product of vectors. i p i+1 The area of ​​triangle p0 is s iThe sum of the areas of all the triangles is the area of ​​the closed shape of the cross section of the material scanned in a single point cloud data session, s = ∑s i .

[0096] The centroid cx of each triangle i =(p0.x+p i .x+p i+1 .x) / 3,cy i =(p0.y+p i .y+p i+1 .y) / 3, the centroid of the closed region 18 is xm=(∑cx i *s i ) / ∑s i ,y m =(∑cy i *s i ) / ∑s i .

[0097] S4, Adjust the material discharge position of the conveyor belt's discharge plate and calculate material loss;

[0098] The area s of the closed region 18 of the material cross-section and the position m(x) of the centroid 19 are obtained from a single scan. m ,y m After that, the following processes can be applied:

[0099] 1) Adjust the material discharge plate actuator according to the change in the material's center of gravity to discharge material at the center of the conveyor belt. Figure 3 As shown, calculate the centroid position m(x) m ,y m The deviation d between the centerline position 3 of the conveyor belt and the centerline position 3 of the conveyor belt. m .

[0100] The maximum deviation between the centroid 19 position of the closed graphic of the material scan section in a single point cloud data and the centerline position of the conveyor belt is d. max (See Figure 3 (as shown by label 23).

[0101] The deviation adjustment threshold value between the centroid position of the closed shape of the material scan section in a single point cloud data and the centerline position of the conveyor belt is d. Δ1 (See Figure 3 (as shown by label 22).

[0102] If the 2D laser scanner 1 is placed above the discharge roller of the chute at the tail of the machine, then:

[0103] If d m ≤d Δ1 This indicates that the material drop point of the unloading plate is within the normal range and no adjustment is needed;

[0104] If dm ≥d max This indicates that the material drop point of the unloading plate is too far off, and the machine should be stopped for inspection.

[0105] If dΔ1<d m <d max This indicates an abnormal material drop point on the unloading plate. This data is fed back to the unloading plate adjustment actuator to perform adjustment. If the unloading plate adjustment actuator is equipped with an electro-hydraulic actuator, automatic adjustment can be achieved; otherwise, manual adjustment can be used. m Adjust to d Δ1 Within the range.

[0106] The material is subjected to t1 to t n By continuously scanning and connecting the calculated centroid positions in chronological order, we obtain... Figure 3 The graph shows the trend of the centroid's change, providing a visualization of the real-time centroid position and the centroid change curve.

[0107] 2) Determine whether material conveying is required based on the start command.

[0108] Combined with the belt running command, the material contour is continuously scanned on the belt conveyor to obtain multiple material scan contours 17. By comparing whether the material scan contours 17 have changed, it can be determined whether the material on the belt conveyor is being conveyed. Similarly, combined with the belt conveyor stop command, it can be determined whether the material on the belt conveyor has stopped being conveyed.

[0109] 3) Calculate the material conveying volume and material loss during the conveying process.

[0110] The single scan time of the 2D laser scanner 1 is t (milliseconds). Given a sufficiently small period t, the material scanning profile 17 can be assumed to remain unchanged between two scans. Within the time length T, calculate the conveying speed v (m / min) of the belt conveyor and the conveyed volume on the belt conveyor:

[0111]

[0112] Two-dimensional laser scanners 1 are installed at the head, middle and tail of the conveyor belt. By calculating the difference in the conveying volume V of the conveyor belt within the time length T, and combining it with the bulk density ρ of the material, the material drop and spillage loss during the conveying process is evaluated. The ratio of the material loss during the conveying process to the rated conveying capacity of time T is calculated to evaluate the material conveying loss rate.

[0113] Example

[0114] Combination Figure 4 As shown, the SICK LMS4111 was used with a start angle of 55 degrees, an end angle of 125 degrees, a scan frequency of 600 Hz, and a step angle of 0.0833 degrees.

[0115] The conveyor belt speed is 250 meters per minute. Therefore, the conveying distance of the material in a single scan is 250 / 60*1000 / 600 = 7 mm. In continuous scanning mode, it can be considered that the real-time data of the material shape is obtained.

[0116] The sensor is installed above the material carrier belt, which transitions from a 45° trough to a 30° trough, on the outside of the protective cover at the material unloading point, according to the S1 installation procedure. The inclined side of the roller is 580mm.

[0117] In the absence of material conveying, a single scan is performed, and after filtering, the result is obtained. Figure 4 The diagram showing the connection of the point cloud of the idler roller profile in the unloaded state, as shown in position (a).

[0118] Taking the linear fitting idler roller model as an example, a certain number of point clouds of the left idler roller are selected, and the linear equation f1 corresponding to the hypotenuse of the idler roller is obtained by the least squares method. Similarly, the linear return line f2 corresponding to the hypotenuse of the idler roller and the linear equation f3 corresponding to the bottom edge of the idler roller can be obtained. By calculating the focus of the linear equations f1 and f3, f2 and f3, the intersection point P1(x) of the two hypotenuses of the idler roller and the bottom edge of the idler roller is obtained. c1 ,y c1 ), P1′(x c1 ′,y c1 ′).

[0119] The equations of the lines f1, f2, f3 and their intersection point P1(x) c1 ,y c1 ), P1′(x c1 ′,y c1 ′) is the mathematical model of the linear fitting idler model corresponding to the 30-degree idler.

[0120] During material conveying, continuous scanning is performed. For a single scan, the data obtained is... Figure 4 Point cloud data containing the outer contours of the material and the outer contours of the idler rollers is shown at position (b).

[0121] Set a threshold h to remove all Figure 4 Point cloud data at position (b) that are less than h away from lines f1, f2, and f3 are used to obtain point cloud data. Figure 4 The point cloud of the outer contour of a single pure material is shown at position (c).

[0122] extract Figure 4Points Point 1 (x, y) and Point 2 (x, y) of the pure material outer contour point cloud shown at position (c) are assigned a number of points N. Data between Point 1 and Point 2 is extracted from the 30-degree straight-line fitted roller mathematical model using the formula (Point 2.x - Point 1.x) / N. This yields the bottom roller point cloud of the single material shape, as shown below. Figure 4 As shown in position (d).

[0123] Will Figure 4 The single pure material outer contour point cloud shown at position (c) and as shown in the figure Figure 4 The point cloud of the bottom roller of the single material shape shown at position (d) is constructed according to Method 1, which uses the material shape contour extraction method based on linear fitting of the roller model to construct the outer contour of the material. Figure 4 The bottom roller structure shown in position (e) encloses the material shape, which is the material shape obtained in a single scan.

[0124] exist Figure 4 Based on the closed material shape of the bottom roller structure shown in position (e), with the lower left intersection point P1(x) c1 ,y c1 Starting from ) Figure 4 The bottom idler structure shown in position (e) is used to close the material shape and cut it into pieces. Figure 4 The multiple triangles shown at position (f) are used to calculate the area and centroid of the material shape in a single scan according to the algorithm for the area and centroid of the material shape in this paper.

[0125] Based on the area and centroid mentioned above, and combined with the material discharge centerline of the conveyor belt, controlling the guide plate is a deviation control measure, which will not be elaborated further.

[0126] Based on the material's area and conveying speed, the volume of the material being conveyed can be obtained. Based on the material's volumetric efficiency and volume, the amount of material conveyed at the current idler position can be easily determined.

[0127] By installing LMS4111 sensors at different locations on the conveyor belt, the difference in conveying volume within the same time period can be calculated, and the material loss during the material conveying process can be easily obtained.

[0128] Those skilled in the art should recognize that the above embodiments are merely illustrative of the present invention and are not intended to limit the present invention. Any variations or modifications to the above embodiments that are within the spirit and essence of the present invention will fall within the scope of the claims of the present invention.

Claims

1. A method for detecting the shape of materials on a belt conveyor using a two-dimensional laser scanner, characterized in that, Includes the following steps: S1, Install a 2D laser scanner; S2, model of idler roller for calibrating conveyor belt machine; S3, Detect the material shape of the conveyor belt and calculate the centroid; S4, Adjust the material discharge position of the conveyor belt's discharge plate and calculate material loss.

2. The method for detecting the shape of materials on a belt conveyor using a two-dimensional laser scanner according to claim 1, characterized in that, The specific steps of S1 are as follows: The two-dimensional laser scanner is positioned directly above the idler roller of the conveyor belt machine; The scanning line plane of the two-dimensional laser scanner is perpendicular to the center line of the belt of the conveyor belt machine and intersects the upper contour line of the idler roller below the belt; When there is no material on the belt, the 0-degree scanning line of the two-dimensional laser scanner is perpendicular to and intersects the midpoint of the idler roller.

3. The method for detecting the shape of materials on a belt conveyor using a two-dimensional laser scanner according to claim 2, characterized in that, Step S2 is as follows: A Cartesian coordinate system is established with the center of the scan line plane of the two-dimensional laser scanner as the 0 point; In the Cartesian coordinate system, the positive direction of the y-axis is vertically downward, and the positive direction of the x-axis is perpendicular to the center line of the belt and away from the center line. When there is no material on the belt and it is stationary, the two-dimensional laser scanner performs a line scan laser scan to acquire idler roller model data.

4. The method for detecting the shape of materials on a belt conveyor using a two-dimensional laser scanner according to claim 3, characterized in that, Step S3 is as follows: In the material conveying state, the two-dimensional laser scanner performs one scan according to the sampling cycle to obtain point cloud data of a single scan. The point cloud data is then filtered and repeated to obtain material contour point cloud data. Combined with the idler roller model, a closed cross-sectional shape of the material scanned by the single point cloud data is constructed. The area of ​​the closed cross-sectional shape of the material scanned by the single point cloud data, the centroid, and the offset between the centroid and the center line of the belt are calculated.

5. The method for detecting the shape of materials on a belt conveyor using a two-dimensional laser scanner according to claim 4, characterized in that, The point cloud data is filtered in the following specific way: Each single scan yields a set of data points a in polar coordinates along the upper outer contour of the idler roller. The set of data points a is then converted into a set of data points b in the Cartesian coordinate system. After performing discrete point filtering and conditional filtering on the set of data points b, a set of data points c is obtained where all points have an angle greater than the angle between the hypotenuse vertex of one side of the idler roller and the x-axis, and less than the angle between the hypotenuse vertex of the other side of the idler roller and the positive x-axis. When there is no material on the belt, the idler model is extracted from the data point set c and used as the reference model for subsequent detection.

6. The method for detecting the shape of materials on a belt conveyor using a two-dimensional laser scanner according to claim 5, characterized in that: Based on the angle of the idler roller and the processing speed of the point cloud data, the methods for extracting the idler roller model specifically include linear fitting of the idler roller model, polynomial fitting of the idler roller model, and direct use of the data point set.

7. The method for detecting the shape of materials on a belt conveyor using a two-dimensional laser scanner according to claim 6, characterized in that, The linear fitting idler roller model specifically includes: For the data point set c, a straight line fitting method is used to obtain the fitted straight line f1 of the hypotenuse of the other side of the idler roller, the fitted straight line f2 of the hypotenuse of one side of the idler roller, and the fitted straight line f3 of the bottom of the idler roller. The coordinates P1′(x) of the intersection point between the hypotenuse of one side of the idler roller and the bottom of the idler roller are then calculated. c1 ′,y c1 The coordinates of the intersection point P1(x′) between the hypotenuse of the other side of the idler roller and the bottom of the idler roller are given. c1 ,y c1 The fitted line f1, the fitted line f2 and the fitted line f3, and the intersection point P1(x) c1 ,y c1 The intersection point P1′(x) c1 ′,y c1 The straight-line fitting model d is formed. In the material conveying state, the two-dimensional laser scanner performs a single scan to obtain point cloud data p1{(x0,y0),(x1,y1),(x2,y2),...,(x n ,y n )}gather.

8. The method for detecting the shape of materials on a belt conveyor using a two-dimensional laser scanner according to claim 6, characterized in that, The polynomial fitting idler roller model specifically includes: For the data point set c, the function f(x) = a n x n +a n-1 x n-1 By fitting +......+a0, we obtain the idler roller model e; For the idler roller model, a quadratic curve is used.

9. The method for detecting the shape of materials on a belt conveyor using a two-dimensional laser scanner according to claim 6, characterized in that, The direct use of the data point set specifically includes: The data point set c is directly used as the idler roller model g.

10. The method for detecting the shape of materials on a belt conveyor using a two-dimensional laser scanner according to claim 7, characterized in that, Material shape contour extraction based on the linear fitting idler roller model specifically includes: Remove all points in the data point set c whose distance to the fitted lines f1 and f2 is less than a set threshold h, thereby obtaining the point cloud dataset p2{(x0′,y0′),(x1′,y1′),(x2′,y2′),...,(x m ′,y m ′)}; The point cloud dataset p2 is combined with the intersection point coordinates P1′(x) c1 ′,y c1 The intersection point coordinates P1(x′) c1 ,y c1 Construct a new point cloud dataset p3{x0′,y0′),(x1′,y1′),(x2′,y2′),...,(x m ′,y m ′), (x c1 y c1 ),(x c1 ′,y c1 Starting from p3{x0′,y0′), following the order of data in the point cloud set, the data points in p3 are sequentially connected using fold lines in a clockwise direction to obtain the region of the closed graphic of the single point cloud data material scan section.

11. The method for detecting the shape of materials on a belt conveyor using a two-dimensional laser scanner according to claim 8, characterized in that, The material shape contour extraction based on the polynomial fitting idler roller model specifically includes: Remove all points from the data set c to fit the polynomial f(x) = a n x n +a n-1 x n-1 Points whose distance from a0 is less than a set threshold h are used to obtain the point cloud dataset p2{(x0′,y0′),(x1′,y1′),(x2′,y2′),...,(x m ′,y m Then extract f(x) = a')}. n x n +a n-1 x n-1 +......+a0, the fitted curve lies between x0′ and x m Point cloud dataset p4{(x) between ' p0 ′,y p0 ′),(x p1 ′,y p1 ′),(x p2 ′,y p2 ′),...,(x pm ′,y pm The point cloud datasets p2 and p4 are combined to form a new point cloud dataset p3{(x0′,y0′),(x1′,y1′),(x2′,y2′),...,(x′)}, where p2 and p4 are combined to form a new point cloud dataset p3{(x0′,y0′),(x1′,y1′),(x2′,y2′),...,(x′)}. m ′,y m ′),(x pm ′,y pm ′),...,(x p0 ′,y p0 ′)}, thereby obtaining the region of the closed graphic of the material scan section of the single point cloud data.

12. The method for detecting the shape of materials on a belt conveyor using a two-dimensional laser scanner according to claim 9, characterized in that, Material shape contour extraction based on the aforementioned direct use of data point sets specifically includes: All point clouds within the overlap threshold h between the data point set c and the idler roller model g are removed to obtain the point cloud dataset p2{(x0′,y0′),(x1′,y1′),(x2′,y2′),...,(x m ′,y m Then extract the idler roller model g between x0′ and x′. m Point cloud dataset p4{(x) between ' p0 ′,y p0 ′),(x p1 ′,y p1 ′),(x p2 ′,y p2 ′),...,(x pm ′,y pm The point cloud datasets p2 and p4 are combined to form a new point cloud dataset p3{(x0′,y0′),(x1′,y1′),(x2′,y2′),...,(x′)}, where p2 and p4 are combined to form a new point cloud dataset p3{(x0′,y0′),(x1′,y1′),(x2′,y2′),...,(x′)}. m ′,y m ′),(x pm ′,y pm ′),...,(x p0 ′,y p0 ′)}, thereby obtaining the region of the closed graphic of the material scan section of the single point cloud data.

13. The method for detecting the shape of materials on a belt conveyor using a two-dimensional laser scanner according to any one of claims 10-12, characterized in that, The specific calculation of the region and area of ​​the closed shape of the material scan section in a single point cloud data is as follows: The point cloud dataset p3 is equivalently represented according to the original point cloud order as p3{p0,p1,p2,...,p...} n }; With p0 as the origin, p i and p i+1 For the other two vertices of the triangle, according to p0p i p i+1 p0 sequentially constructs vector triangles, dividing the region of the closed graphic of the single point cloud data material scan section into n-2 vector triangles; Calculate p0p using the cross product of vectors. i p i+1 The area of ​​triangle p0 is s i The sum of the areas of all the triangles is the area s of the closed shape of the cross-section of the material scanned in a single point cloud data session, which is s = ∑s i . The centroid cx of each triangle i =(p0.x+p i .x+p i+1 .x) / 3,cy i =(p0.y+p i .y+p i+1 .y) / 3, the centroid of the closed region 18 is xm=(∑cx i *s i ) / ∑s i ,y m =(∑cy i *s i ) / ∑s i .

14. The method for detecting the shape of materials on a belt conveyor using a two-dimensional laser scanner according to claim 13, characterized in that, The specific steps in step S4, such as adjusting the material discharge position of the tape machine's feeding plate, include: Calculate the position of the centroid m(x) m ,y m The deviation d between the centerline position of the conveyor belt and the centerline position of the conveyor belt machine m ; The maximum deviation between the centroid of the closed section of the material scanned from the single point cloud data and the centerline of the conveyor belt is d. max ; The deviation adjustment threshold value between the centroid position of the closed shape of the material scan section of the single point cloud data and the centerline position of the conveyor belt is d. Δ1 ; If d m ≤d Δ1 This indicates that the material drop point of the pouring plate is within the normal range and no adjustment is required; If d m ≥d max This indicates that the material drop point of the discharge plate is too far off, and the machine should be stopped for inspection. If d Δ1 <d m <d max This indicates that the material drop point of the discharge plate is abnormal.

15. The method for detecting the shape of materials on a belt conveyor using a two-dimensional laser scanner according to claim 13, characterized in that, The calculation of material loss in step S4 specifically includes: The scanning time of the two-dimensional laser scanner is t. Within the time length T, calculate the conveying speed v of the tape machine and the volume conveyed on the tape machine: A two-dimensional laser scanner is installed at the head, middle and tail of the conveyor belt. By calculating the difference in the conveying volume V of the conveyor belt within a time length T, and combining it with the bulk density ρ of the material, the material drop and spillage loss during the conveying process is evaluated. The ratio of the material loss during the conveying process to the rated conveying capacity of time T is calculated to evaluate the material conveying loss rate.

Images