Unattended bucket wheel machine path optimization method based on industrial vision

Abstract

The application relates to the field of industrial vision technology, and particularly discloses an unmanned attended bucket wheel machine taking path optimization method based on industrial vision, a stockpile three-dimensional surface model is established through industrial vision, geometric cross-sectional areas at different positions along a rotary taking path are calculated, and a rotary angular velocity is inversely deduced from a reference volume flow rate; then, a speed saturation section is identified in combination with upper limits of rotary mechanism speed and acceleration, a flow shortage generated in the section is calculated, and finally, compensation adjustment is carried out in a front stable taking path section to form a new rotary angular velocity instruction and a bucket wheel cutting parameter.

Description

Technical Field

[0001] This invention relates to the field of industrial vision technology, and more specifically to a method for optimizing the material handling path of an unattended bucket wheel excavator based on industrial vision. Background Technology

[0002] Bucket wheel excavators are widely used in continuous material handling operations in bulk material stockpiles such as coal yards. Unmanned bucket wheel excavators typically need to generate a rotary material handling path based on the surface morphology of the stockpile, the target material handling flow rate, and the equipment motion parameters.

[0003] When the bucket wheel excavator is taking material from the main area of ​​the material pile, the cutting cross-sectional area between the bucket wheel and the material pile changes relatively smoothly, and the material flow rate on the rotation path is relatively easy to maintain continuously. However, in the edge area of ​​the material pile, the thin material area, or the cutting-in and cutting-out area of ​​the rotation path, the geometric cross-sectional area between the bucket wheel and the material pile will change significantly within a shorter path range, and the theoretical rotational angular velocity calculated according to the constant flow rate is prone to local sudden increases.

[0004] The rotary mechanism is limited by the output speed and acceleration capabilities. The theoretical rotational angular velocity is restricted to the allowable range during actual operation, resulting in a lower actual material flow rate within the speed saturation zone compared to the baseline flow rate. This leads to localized flow dips on the downstream conveyor belt. For unattended material handling processes, these localized flow dips affect the stability of the material flow during continuous material handling and increase the complexity of subsequent path adjustments and material handling rhythm control. Summary of the Invention

[0005] The purpose of this invention is to provide a method for optimizing the material handling path of an unattended bucket wheel excavator based on industrial vision, thereby solving the above-mentioned technical problems.

[0006] The objective of this invention can be achieved through the following technical solutions: The method for optimizing the material handling path of an unattended bucket wheel excavator based on industrial vision includes the following steps: The point cloud data of the material pile surface is collected by an industrial vision system, and a three-dimensional surface model of the material pile is established based on the point cloud data of the material pile surface. Obtain the baseline volumetric flow rate, the upper limit of the output speed of the rotary mechanism, and the upper limit of the acceleration; Calculate the geometric cross-sectional area point by point along the preset rotary material handling path, and calculate the theoretical rotational angular velocity based on the reference volumetric flow rate, geometric cross-sectional area and rotational radius; Based on the theoretical rotational angular velocity and the upper limit values ​​of velocity and acceleration, the velocity saturation section on the rotational material handling path is identified; The actual angular velocity curve is generated based on the upper limit values ​​of velocity and acceleration, and the flow deficit in the velocity saturation section is calculated by integration. Before the speed saturation zone, a stable material feeding path section is selected, the geometric cross-sectional area is increased and the rotational angular velocity is reduced, and the adjusted rotational angular velocity command sequence and bucket wheel cutting parameters are generated.

[0007] As a further aspect of the present invention: the process of establishing a three-dimensional surface model of the stockpile is as follows: Multiple two-dimensional images of the material pile are acquired using a variety of fixedly installed image acquisition devices; A stereo vision matching algorithm is used to process multi-view two-dimensional images to generate an initial dense point cloud. The initial dense point cloud is then subjected to statistical outlier removal filtering to obtain a filtered point cloud. The filtered point cloud is triangulated to generate a continuous surface mesh composed of triangular facets, which is then used as the three-dimensional surface model of the material pile.

[0008] As a further aspect of the present invention: the process of obtaining the reference volumetric flow rate and the upper limit values ​​of the output speed and acceleration of the rotary mechanism is as follows: The reference material flow rate is received from the upper-level process control system of the bucket wheel excavator, the material bulk density is read from the material yard database, and the reference material flow rate is divided by the material bulk density to obtain the reference volumetric flow rate. Read the maximum output speed, maximum output torque, transmission ratio, and cantilever moment of inertia of the slewing motor from the parameters of the slewing mechanism drive device of the bucket wheel excavator; The upper limit of the angular velocity at the output end of the rotary mechanism is calculated based on the maximum output speed and transmission ratio of the rotary motor and used as the upper speed limit; The upper limit of the angular acceleration at the output end of the rotary mechanism is calculated based on the maximum output torque of the rotary motor, the transmission ratio, and the moment of inertia of the cantilever.

[0009] As a further aspect of the present invention: the process of calculating the theoretical angular velocity is as follows: Transform the three-dimensional surface model of the material pile to the coordinate system of the bucket wheel excavator cantilever kinematics model; The preset rotary material handling path is discretized into multiple path sampling points. Based on the cantilever rotation angle, cantilever pitch angle, cantilever length and bucket wheel installation offset corresponding to the path sampling point read from the cantilever kinematic model of the bucket wheel excavator, the coordinates of the bucket wheel center of the path sampling point are calculated. The horizontal distance from the center of the bucket wheel to the center of rotation of the cantilever is taken as the rotation radius corresponding to the path sampling point; The radial direction of the cantilever is determined based on the cantilever rotation angle. The plane that passes through the center coordinates of the bucket wheel and is jointly defined by the radial and vertical directions of the cantilever is taken as the cutting section plane corresponding to the path sampling point. Traverse the triangular facets in the 3D surface model of the material pile, calculate the intersection points of the edges of the triangular facets with the cutting section plane, connect the two intersection points on the same triangular facet to form the facet intersection line, and connect all the facet intersection lines according to the adjacency relationship of the endpoints of the facet intersection lines to obtain the material pile section polygonal line; Obtain the outer radius of the bucket teeth, and generate a circular boundary in the cutting section plane based on the coordinates of the bucket wheel center and the outer radius of the bucket teeth as the cutting envelope of the bucket wheel; Obtain the intersection point of the material pile section polygon and the bucket wheel cutting envelope. Select the material pile section polygon located inside the bucket wheel cutting envelope. Connect the selected material pile section polygon with the bucket wheel cutting envelope between the intersection points to form a closed cutting section profile. The area of ​​the closed cutting section profile is calculated as the geometric cross-sectional area. The cutting linear velocity is obtained by dividing the reference volumetric flow rate by the geometric cross-sectional area. The cutting linear velocity is then divided by the radius of gyration corresponding to the path sampling point to obtain the theoretical angular velocity of gyration corresponding to the path sampling point.

[0010] As a further aspect of the present invention: the process of identifying the velocity saturation zone is as follows: Arrange the theoretical rotational angular velocities corresponding to the sampling points of each path in the order of the preset rotary material handling path to obtain the angular velocity sequence; Calculate the absolute difference between the i-th theoretical angular velocity and the previous theoretical angular velocity in the angular velocity sequence, and divide the absolute difference by the preset motion time to obtain the theoretical angular acceleration; If the theoretical angular acceleration is greater than the upper limit of acceleration, then the i-th theoretical angular velocity is marked as an abnormal velocity; The theoretical angular velocity exceeding the speed limit is also marked as an abnormal velocity. The path segment consisting of path sampling points corresponding to continuous abnormal speeds is marked as a speed saturation segment.

[0011] As a further aspect of the present invention: the process of calculating the traffic deficit is as follows: Sort the abnormal speeds according to the order of the corresponding path sampling points; For a single abnormal speed, replace it with the smaller value between the speed limit and the abnormal speed. After the replacement is completed, all elements in the sort are recorded as one type of speed. Obtain the theoretical turning angular velocity C1 corresponding to the previous path sampling point adjacent to the starting point of the velocity saturation section, and add C1 as the first digit of the sorting; Calculate the product S of the upper limit of acceleration and the preset motion time; Starting with the first type of speed A1 in the sort, add the product S to speed A1 to get the maximum value MAX of the second type of speed A2 in the sort, and subtract the product S from speed A1 to get the minimum value MIN of speed A2. If A2 > MAX, then let A2 = MAX; if A2 < MIN, then let A2 = MIN; if MIN ≤ A2 ≤ MAX, then keep A2 unchanged. Use A2 as A1 in the next iteration, and use A3, the third-ranked velocity in the sort, as A2 in the next iteration. Repeat the calculation of MAX and MIN and adjust A2 until the last velocity in the sort is adjusted. After the adjustment is completed, all elements of type I except C1 will be recorded as type II velocities; Multiply the second type of velocity, turning radius and geometric cross-sectional area to obtain the actual material volume flow rate corresponding to the path sampling point. Subtract the actual material volume flow rate from the benchmark volume flow rate to obtain the volume flow rate difference. If the volume flow rate difference is less than zero, set the volume flow rate difference to zero. The volumetric flow rate difference corresponding to each path sampling point within the velocity saturation section is multiplied by the preset motion time, and the flow rate deficit is obtained by summing all the products.

[0012] As a further aspect of the present invention: the process of generating the adjusted angular velocity command sequence and bucket wheel cutting parameters is as follows: The rated volume conveying capacity of the downstream conveyor belt is received from the upstream process control system of the bucket wheel excavator. Before the start of the speed saturation zone, a continuous path sampling point that is not marked as the speed saturation zone is selected in reverse along the preset rotary material picking path, and the path segment corresponding to the continuous path sampling point is taken as the stable material picking path segment. For the path sampling points within the stable material feeding path segment, the cutting section profile, geometric cross-sectional area, radius of gyration, and theoretical angular velocity are obtained. The cutting section profile is translated along the cutting feed direction of the bucket wheel, and the area of ​​the translated cutting section profile is calculated. The area of ​​the translated cutting section profile is used as the adjusted geometric cross-sectional area. The theoretical rotational angular velocity corresponding to the path sampling point in the stable material picking path section is reduced by a preset reduction ratio to obtain the adjusted rotational angular velocity. Multiply the adjusted geometric cross-sectional area, the radius of gyration, and the adjusted angular velocity of gyration to obtain the adjusted actual material collection volume flow rate corresponding to the path sampling point within the stable material collection path segment; If the adjusted actual material intake volume flow rate is greater than the reference volume flow rate but less than or equal to the rated volume conveying capacity, then the difference between the adjusted actual material intake volume flow rate and the reference volume flow rate is recorded as the compensation volume flow rate difference. The difference in compensated volumetric flow rate at each sampling point within the stable material intake path segment is multiplied by the preset motion time, and the sum of all products is obtained to obtain the compensated flow rate increment. Adjust the translation amount of the cutting section profile and the preset reduction ratio until the compensation flow increment is equal to the flow deficit; The translation of the cutting section profile, the adjusted geometric cross-sectional area, and the adjusted rotational angular velocity when the compensated flow increment is equal to the flow deficit are used as the cutting parameters of the bucket wheel. The adjusted rotational angular velocity within the stable material handling path segment, the second-class velocity within the velocity saturation segment, and the theoretical rotational angular velocity after the velocity saturation segment are spliced ​​together in the order of the preset rotational material handling path to form the adjusted rotational angular velocity command sequence.

[0013] The beneficial effects of this invention compared to the prior art are as follows: In this invention, unexecutable speed sections in the rotary material handling path are identified during the path planning stage, the material flow deficit corresponding to speed saturation sections is quantified, and the preceding stable material handling path sections compensate for subsequent flow dips, reducing the fluctuation amplitude of material flow received by the downstream conveyor belt. The adjusted rotary angular velocity command sequence matches the speed and acceleration boundaries at the output end of the rotary mechanism, improving the executability of path commands and reducing the sense of local material handling interruption in edge and thin material areas. The bucket wheel excavator's material handling rhythm is more continuous in unattended operation, the instantaneous low flow range of the conveyor belt is reduced, and the material handling path changes from simply calculating speed based on geometric cross-sectional area to a path scheduling result that simultaneously considers material flow deficit and mechanism execution boundaries, thereby improving the material flow stability and path execution consistency during continuous material handling by the unattended bucket wheel excavator. Attached Figure Description

[0014] The invention will now be further described with reference to the accompanying drawings.

[0015] Figure 1 This is a flowchart illustrating the unattended bucket wheel excavator material handling path optimization method based on industrial vision according to the present invention. Figure 2 This is a schematic diagram of the process for generating the adjusted slewing angular velocity command sequence and bucket wheel cutting parameters according to the present invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] Please see Figures 1-2 As shown, this invention is a method for optimizing the material handling path of an unattended bucket wheel excavator based on industrial vision, including the following steps: The point cloud data of the material pile surface is collected by an industrial vision system, and a three-dimensional surface model of the material pile is established based on the point cloud data of the material pile surface. In a preferred embodiment of the present invention, the process of establishing a three-dimensional surface model of the material pile is as follows: Multiple two-dimensional images of the material pile are acquired using a variety of fixedly installed image acquisition devices; A stereo vision matching algorithm is used to process multi-view two-dimensional images to generate an initial dense point cloud. The initial dense point cloud is then subjected to statistical outlier removal filtering to obtain a filtered point cloud. The filtered point cloud is triangulated to generate a continuous surface mesh composed of triangular facets, which is then used as the three-dimensional surface model of the material pile.

[0018] Obtain the baseline volumetric flow rate, the upper limit of the output speed of the rotary mechanism, and the upper limit of the acceleration; In another preferred embodiment of the present invention, the process of obtaining the reference volumetric flow rate and the upper limit values ​​of the output speed and acceleration of the rotary mechanism is as follows: The reference material handling flow rate corresponding to the current material handling task is read from the upper-level process control system of the bucket wheel excavator. The reference material handling flow rate is stored in the form of mass flow rate. Then, based on the material pile identifier corresponding to the current material handling task, the material bulk density corresponding to the current material pile is read from the material yard database. After reading the reference material handling flow rate and the material bulk density, the mass units of the reference material handling flow rate and the material bulk density are unified, and the time units of the reference material handling flow rate and the time units used for subsequent rotation path calculation are unified. Then, the reference material handling flow rate with unified units is divided by the material bulk density to obtain the material volume per unit time. The material volume per unit time is recorded as the reference volumetric flow rate.

[0019] The maximum output speed, maximum output torque, transmission ratio, and cantilever moment of inertia of the slewing mechanism drive unit of the bucket wheel excavator are read from the parameters. The maximum output speed of the slewing motor is the maximum angular velocity allowed to be reached by the output shaft; the maximum output torque is the maximum torque allowed to be output by the output shaft; the transmission ratio is the ratio between the angular velocity of the output shaft and the angular velocity at the output end of the slewing mechanism; the cantilever moment of inertia is the moment of inertia corresponding to the rotation of the cantilever, bucket wheel, and follower components mounted on the cantilever around the cantilever's rotation center. If the maximum output speed of the slewing motor is stored in revolutions per minute (RPM), it is first converted to radians per second according to the speed unit conversion rules to obtain the maximum output angular velocity of the slewing motor; if the maximum output speed is already stored in radians per second, it is directly recorded as the maximum output angular velocity of the slewing motor. Divide the maximum output angular velocity of the rotary motor by the transmission ratio to obtain the upper limit value of the output angular velocity of the rotary mechanism, and record the upper limit value of the output angular velocity of the rotary mechanism as the speed upper limit value.

[0020] When calculating the upper limit of the angular acceleration at the output end of the slewing mechanism, first multiply the maximum output torque of the slewing motor by the transmission ratio to obtain the maximum driving torque at the output end of the slewing mechanism; then divide the maximum driving torque at the output end of the slewing mechanism by the moment of inertia of the cantilever to obtain the upper limit of the angular acceleration at the output end of the slewing mechanism, and record this upper limit of the angular acceleration at the output end of the slewing mechanism as the upper limit of acceleration. The upper limit of velocity corresponds to the maximum executable angular velocity of the cantilever during slewing motion, and the upper limit of acceleration corresponds to the maximum executable angular acceleration of the cantilever during slewing motion.

[0021] Calculate the geometric cross-sectional area point by point along the preset rotary material handling path, and calculate the theoretical rotational angular velocity based on the reference volumetric flow rate, geometric cross-sectional area and rotational radius; In another preferred embodiment of the present invention, the process of calculating the theoretical angular velocity is as follows: The transformation relationship between the coordinate system of the 3D surface model of the material pile and the coordinate system of the kinematic model of the bucket wheel excavator cantilever is read. The transformation relationship includes translation and rotation. The homogeneous coordinate transformation method is used to transform the vertex coordinates of each triangular facet in the 3D surface model of the material pile, and the transformed vertex coordinates are obtained. The set of triangular facets formed by the transformed vertex coordinates is denoted as the transformed 3D surface model of the material pile.

[0022] The 3D surface model of the material pile and the kinematic model of the bucket wheel excavator cantilever are derived from different coordinate references. The material pile coordinates obtained from point cloud reconstruction typically correspond to the coordinates of the image acquisition device or the material yard measurement, while the bucket wheel center, cantilever rotation angle, and cantilever pitch angle correspond to the bucket wheel excavator's own motion coordinates. In 3D space, the same object can be expressed as having the same spatial position in different coordinate systems through translation and rotation relationships. The homogeneous coordinate transformation method, an existing coordinate transformation method in 3D geometry, can transform the vertices of the triangular facets from the material pile coordinates to the coordinate system of the bucket wheel excavator cantilever kinematic model. After the coordinate transformation, the material pile surface and the bucket wheel's motion attitude are under the same geometric reference, allowing for subsequent calculations of the intersection relationship between the cutting section plane and the material pile triangular facets in the same space, and ensuring that the calculation of the geometric cross-sectional area does not result in coordinate reference confusion.

[0023] The system reads rotation angle samples arranged in the order of material handling from the preset rotation path, and records each rotation angle sample as a path sampling point. Each path sampling point represents a rotational posture position of the cantilever within the rotation path. For any path sampling point, the system reads the corresponding cantilever rotation angle, cantilever pitch angle, cantilever length, and bucket wheel installation offset from the bucket wheel excavator cantilever kinematic model. The cantilever rotation angle determines the radial direction of the cantilever in the horizontal plane, the cantilever pitch angle determines the tilt state of the cantilever in the vertical direction, the cantilever length determines the position of the cantilever tip relative to the cantilever rotation center, and the bucket wheel installation offset represents the positional offset of the bucket wheel center relative to the cantilever tip reference point.

[0024] The preset rotary material handling path essentially corresponds to the working trajectory formed by the continuous rotation of the cantilever around the rotation center. This continuous trajectory cannot be directly calculated point-by-point; it needs to be discretized into path sampling points arranged according to the working sequence. Each path sampling point represents a posture position of the cantilever within the rotary path. The cantilever rotation angle gives the direction in the horizontal plane, the cantilever pitch angle gives the tilt state in the vertical direction, the cantilever length gives the distance of the cantilever front end relative to the rotation center, and the bucket wheel installation offset gives the positional relationship between the bucket wheel center and the reference point at the cantilever front end. In rigid body kinematics, given the connecting rod length, posture angle, and installation offset, the spatial coordinates of the end mounting point can be determined. Therefore, the bucket wheel center coordinates can be calculated from the cantilever posture corresponding to the path sampling point. Once the bucket wheel center coordinates are clear, the cutting position of the bucket wheel at that path sampling point is determined, and the geometric cross-sectional area calculation has a clear spatial center.

[0025] The radial direction of the cantilever is determined based on its rotation angle, and the cantilever length is decomposed into a horizontal projected length and a vertical height component based on its pitch angle. Starting from the cantilever rotation center, the horizontal projected length is moved radially along the cantilever, and the vertical height component is superimposed vertically to obtain the coordinates of the reference point at the cantilever front end. Then, according to the cantilever rotation and pitch angles, the bucket wheel installation offset is converted from the local coordinate relationship of the cantilever to an offset quantity in the coordinate system of the bucket wheel excavator cantilever kinematic model. This offset quantity is superimposed onto the coordinates of the reference point at the cantilever front end to obtain the coordinates of the bucket wheel center corresponding to the path sampling point.

[0026] The slewing radius is the horizontal distance from the center of the bucket wheel to the center of rotation of the cantilever. The cantilever rotation occurs in the horizontal direction, and the tangential linear velocity of the bucket wheel along the rotation path is determined by both the slewing radius and the angular velocity. The distance between the projected center of the bucket wheel onto the horizontal plane and the center of rotation of the cantilever represents the slewing radius corresponding to the movement of the bucket wheel center around the center of rotation. The cutting section plane is selected as a plane passing through the coordinates of the bucket wheel center and jointly defined by the radial and vertical directions of the cantilever. This plane is equivalent to making a vertical cut along the current cantilever direction on the stockpile and the bucket wheel. When the bucket wheel cuts into the stockpile at the sampling point of this path, the stockpile profile and the outer edge of the bucket teeth form a two-dimensional cross-sectional relationship within this cutting plane, and the three-dimensional cutting problem is transformed into a two-dimensional cross-sectional area calculation problem.

[0027] In the coordinate system of the bucket wheel excavator cantilever kinematic model, read the coordinates of the cantilever rotation center and the bucket wheel center. Project the bucket wheel center coordinates onto the horizontal plane, and calculate the distance between the projected bucket wheel center coordinates and the cantilever rotation center in the horizontal plane. Use this distance as the rotation radius corresponding to the path sampling point. Determine the cantilever radial direction based on the cantilever rotation angle, and use the plane passing through the bucket wheel center coordinates and simultaneously containing both the cantilever radial and vertical directions as the cutting section plane corresponding to the path sampling point.

[0028] The process iterates through the triangular faces in the transformed 3D surface model of the stockpile. For each individual triangular facet, the positional relationship of its three vertices relative to the cutting section plane is determined. If the two endpoints of one side of a triangular facet are located on opposite sides of the cutting section plane, the intersection point of that side with the cutting section plane is calculated. If the endpoint of one side of a triangular facet is located on the cutting section plane, the endpoint on the cutting section plane is taken as the intersection point. For the same triangular facet, when two intersection points are obtained, the two intersection points are connected to form a facet intersection line. When the boundary line segment of the triangular facet is located within the cutting section plane, the boundary line segment within the cutting section plane is taken as the facet intersection line. After traversing all triangular faces, all facet intersection lines are connected according to the coordinate relationship of the endpoints of the facet intersection lines to obtain the stockpile section polygonal line corresponding to the path sampling point.

[0029] The 3D surface model of the material pile is composed of triangular facets. When the cutting section plane passes through the surface of the material pile, it will intersect with some of the triangular facets. In 3D computational geometry, when a plane intersects with a triangular facet, the intersection result is usually represented as a line segment, endpoint, or coincident boundary. By determining the positional relationship of each vertex of the triangular facet relative to the cutting section plane, the intersection points of the triangular facet edges and the cutting section plane can be determined; connecting two intersection points on the same triangular facet forms a facet intersection line. Connecting all facet intersection lines according to the adjacency relationship of their endpoints forms a section polygon of the material pile within the current cutting section plane. This polygon represents the boundary of the material pile surface within the cutting plane under the current cantilever direction.

[0030] The outer radius of the bucket teeth is read from the structural parameters of the bucket wheel excavator. The outer radius of the bucket teeth is the distance between the center of the bucket wheel and the outermost cutting edge of the teeth. Within the cutting section plane, a circular boundary is generated using the bucket wheel center coordinates and the outer radius of the bucket teeth. This circular boundary is denoted as the bucket wheel cutting envelope. The bucket wheel cutting envelope represents the boundary swept by the outer edge of the bucket teeth within the section corresponding to the path sampling point.

[0031] The cutting range of the bucket wheel is determined by the outer edge of the bucket teeth, and the radius of the outer edge of the bucket teeth represents the distance between the center of the bucket wheel and the outermost cutting end of the bucket teeth. Within the cutting section plane, the coordinates of the bucket wheel center and the radius of the outer edge of the bucket teeth define a circular boundary, which represents the outer boundary swept by the outer edge of the bucket teeth within this cutting section. The actual material pile section that the bucket wheel can cut must be within the coverage area of ​​the circular boundary of the outer edge of the bucket teeth. The relative positional relationship between the circular boundary and the broken line of the material pile section expresses the contact range between the bucket wheel and the material pile. After denoteing this circular boundary as the bucket wheel cutting envelope, the bucket wheel geometry is transformed into a calculable two-dimensional boundary.

[0032] Calculate the intersection point between the broken line of the stockpile cross section and the cutting envelope of the bucket wheel. For each broken line segment in the broken line of the stockpile cross section, calculate the distance from the endpoint of the broken line segment to the coordinates of the bucket wheel center. If the distance from the endpoint of the broken line segment to the coordinates of the bucket wheel center is less than or equal to the outer radius of the bucket tooth, the endpoint is recorded as the endpoint located inside the cutting envelope of the bucket wheel. If both endpoints of the broken line segment are located inside the cutting envelope of the bucket wheel, the broken line segment is retained. If one endpoint of the broken line segment is located inside the cutting envelope of the bucket wheel and the other endpoint is located outside the cutting envelope of the bucket wheel, the broken line segment is divided at the intersection point of the broken line segment and the cutting envelope of the bucket wheel, and the portion of the broken line segment located inside the cutting envelope of the bucket wheel is retained. The retained continuous broken line segment of the stockpile cross section is taken as the inner broken line of the stockpile cross section.

[0033] In the broken line of the material pile section, only the portion falling inside the bucket wheel's cutting envelope corresponds to the material that the bucket wheel can contact in its current posture. The distance from the endpoint of the broken line segment to the bucket wheel's center coordinates determines whether the endpoint is inside or outside the bucket wheel's cutting envelope. When the broken line segment crosses the bucket wheel's cutting envelope, the intersection of the broken line segment and the bucket wheel's cutting envelope divides the broken line segment into an inner and an outer portion. After retaining the inner broken line of the material pile section, connecting the circular arc segments of the bucket wheel's cutting envelope between the intersection points forms a closed cutting section profile. The closed profile corresponds to the cross-sectional area jointly defined by the outer edge of the bucket wheel and the surface of the material pile; the area of ​​this area represents the geometric cross-sectional area of ​​the bucket wheel cutting the material pile at the current path sampling point.

[0034] For each segment of the inner material pile cross-section polygon, read the two endpoints of the polygon. If the two endpoints are located on the bucket wheel cutting envelope, select the arc segment in the bucket wheel cutting envelope that lies between the two endpoints and forms the area enclosed by the inner material pile cross-section polygon. Connect the inner material pile cross-section polygon with the arc segment to form a closed cutting cross-section profile. If there are multiple segments of the inner material pile cross-section polygon, form multiple closed cutting cross-section profiles respectively. Discretize the arc segments in the closed cutting cross-section profile into arc points according to a preset arc length interval. Arrange the arc points and the polygon points on the inner material pile cross-section polygon in a looping order to obtain a sequence of closed points. Calculate the area enclosed by the sequence of closed points using the shoelace formula, and record the area value as the cross-sectional area of ​​a single closed cutting cross-section profile. Sum the cross-sectional areas of all closed cutting cross-section profiles corresponding to the path sampling points to obtain the geometric cross-sectional area corresponding to the path sampling points.

[0035] The closed cutting section profile is enclosed by the inner material pile section polygonal line and the circular arc segment of the bucket wheel cutting envelope. The circular arc segment cannot be directly calculated using the polygonal line area; it needs to be discretized into arc points at preset arc length intervals. The arc points and polygonal line points are arranged in a looping order to form a closed point sequence. The shoelace formula, an existing method for polygon area calculation, can obtain the area of ​​the enclosed region from the closed point sequence. If multiple separate closed cutting section profiles exist for the same path sampling point, the area of ​​each closed cutting section profile is calculated separately and then summed to obtain the total geometric cross-sectional area corresponding to that path sampling point. This geometric cross-sectional area incorporates the bucket wheel size, material pile profile, and current cantilever posture, and can be used as a cross-sectional quantity in the calculation of the material handling volumetric flow rate.

[0036] After obtaining the geometric cross-sectional area corresponding to the path sampling point, the reference volumetric flow rate is divided by the geometric cross-sectional area to obtain the cutting linear velocity corresponding to the path sampling point. The cutting linear velocity represents the linear velocity in the tangential direction of rotation when the bucket wheel passes the path sampling point along the preset rotary material handling path. The cutting linear velocity is then divided by the rotation radius corresponding to the path sampling point to obtain the theoretical rotational angular velocity corresponding to the path sampling point. The above process is repeated for each path sampling point according to the operation sequence of the preset rotary material handling path to obtain the theoretical rotational angular velocity corresponding to each path sampling point.

[0037] Volumetric flow rate equals the product of cross-sectional area and linear velocity. The baseline volumetric flow rate represents the volume of material to be cut per unit time, while the geometric cross-sectional area represents the cross-sectional area that the bucket wheel can cut at the current path sampling point. Dividing the baseline volumetric flow rate by the geometric cross-sectional area yields the cutting linear velocity required to maintain the baseline volumetric flow rate. In rotary motion, the cutting linear velocity equals the product of the rotational angular velocity and the radius of rotation. Dividing the cutting linear velocity by the radius of rotation corresponding to the path sampling point yields the theoretical rotational angular velocity corresponding to that sampling point. After repeating the calculation for all path sampling points according to the preset rotary material handling path sequence, a sequence of theoretical rotational angular velocities corresponding to the geometric changes of the material pile is formed.

[0038] Based on the theoretical rotational angular velocity and the upper limit values ​​of velocity and acceleration, the velocity saturation section on the rotational material handling path is identified; In a preferred embodiment of the present invention, the process of identifying the velocity saturation zone is as follows: Arrange the theoretical rotational angular velocities corresponding to the sampling points of each path in the order of the preset rotary material handling path to obtain the angular velocity sequence; Calculate the absolute difference between the i-th theoretical angular velocity and the previous theoretical angular velocity in the angular velocity sequence, and divide the absolute difference by the preset motion time to obtain the theoretical angular acceleration; If the theoretical angular acceleration is greater than the upper limit of acceleration, then the i-th theoretical angular velocity is marked as an abnormal velocity; The theoretical angular velocity exceeding the speed limit is also marked as an abnormal velocity. The path segment consisting of path sampling points corresponding to continuous abnormal speeds is marked as a speed saturation segment.

[0039] The actual angular velocity curve is generated based on the upper limit values ​​of velocity and acceleration, and the flow deficit in the velocity saturation section is calculated by integration. In another preferred embodiment of the present invention, the process of calculating the traffic deficit is as follows: Sort the abnormal speeds according to the order of the corresponding path sampling points; For a single abnormal speed, replace it with the smaller value between the speed limit and the abnormal speed. After the replacement is completed, all elements in the sort are recorded as one type of speed. Obtain the theoretical turning angular velocity C1 corresponding to the previous path sampling point adjacent to the starting point of the velocity saturation section, and add C1 as the first digit of the sorting; Calculate the product S of the upper limit of acceleration and the preset motion time; Starting with the first type of speed A1 in the sort, add the product S to speed A1 to get the maximum value MAX of the second type of speed A2 in the sort, and subtract the product S from speed A1 to get the minimum value MIN of speed A2. If A2 > MAX, then let A2 = MAX; if A2 < MIN, then let A2 = MIN; if MIN ≤ A2 ≤ MAX, then keep A2 unchanged. Use A2 as A1 in the next iteration, and use A3, the third-ranked velocity in the sort, as A2 in the next iteration. Repeat the calculation of MAX and MIN and adjust A2 until the last velocity in the sort is adjusted. After the adjustment is completed, all elements of type I except C1 will be recorded as type II velocities; Multiply the second type of velocity, turning radius and geometric cross-sectional area to obtain the actual material volume flow rate corresponding to the path sampling point. Subtract the actual material volume flow rate from the benchmark volume flow rate to obtain the volume flow rate difference. If the volume flow rate difference is less than zero, set the volume flow rate difference to zero. The volumetric flow rate difference corresponding to each path sampling point within the velocity saturation section is multiplied by the preset motion time, and the flow rate deficit is obtained by summing all the products.

[0040] Before the speed saturation zone, a stable material feeding path section is selected, the geometric cross-sectional area is increased and the rotational angular velocity is reduced, and the adjusted rotational angular velocity command sequence and bucket wheel cutting parameters are generated.

[0041] In a preferred embodiment of the present invention, the process of generating the adjusted angular velocity command sequence and bucket wheel cutting parameters is as follows: The rated volume conveying capacity is received from the upstream process control system of the bucket wheel excavator from the downstream conveyor belt. When the rated volume conveying capacity is stored in the form of volumetric flow rate, it is read directly; when the rated volume conveying capacity is stored in the form of mass flow rate, the material bulk density is read from the material yard database, and the mass flow rate corresponding to the rated volume conveying capacity is divided by the material bulk density to obtain the rated volume conveying capacity in the form of volumetric flow rate.

[0042] Before the start of the speed saturation section, path sampling points are read sequentially in the opposite direction of the preset rotary material handling path. When a path sampling point not marked as a speed saturation section is read, it is added to the candidate path sampling point set; when a path sampling point marked as a speed saturation section is read, reading stops. The path sampling points in the candidate path sampling point set are arranged consecutively according to the order in the preset rotary material handling path. The candidate path sampling point adjacent to the start of the speed saturation section is selected first as the initial path sampling point for the stable material handling path section, and then candidate path sampling points are added forward based on the comparison results of the compensation flow increment and flow deficit.

[0043] For a single path sampling point within a stable material handling path segment, read the corresponding cutting section plane, material pile section polygon, bucket wheel cutting envelope, geometric cross-sectional area, radius of gyration, and theoretical angular velocity. Read the polygon point in the material pile section closest to the bucket wheel center coordinates, and use the direction from the bucket wheel center coordinates to this polygon point as the bucket wheel cutting feed direction. The bucket wheel cutting feed direction lies within the cutting section plane and points towards the side of the material pile closest to the bucket wheel in the current section.

[0044] First, set the initial translation amount and initial reduction ratio. Based on the initial translation amount, move the bucket wheel cutting envelope along the bucket wheel cutting feed direction to obtain the translated bucket wheel cutting envelope. Calculate the intersection point between the material pile section polygon and the translated bucket wheel cutting envelope. Select the material pile section polygon located inside the translated bucket wheel cutting envelope. Connect the selected material pile section polygon with the translated bucket wheel cutting envelope between the intersection points to form the adjusted closed cutting section profile. Calculate the area of ​​the adjusted closed cutting section profile and record this area as the adjusted geometric cross-sectional area. Note that here, only the bucket wheel cutting envelope is translated, not the original closed cutting section profile as a whole.

[0045] The adjusted angular velocity is obtained by reducing the theoretical angular velocity corresponding to the path sampling point according to the initial reduction ratio. The adjusted geometric cross-sectional area, angular radius, and angular velocity are multiplied to obtain the adjusted actual material intake volumetric flow rate corresponding to the path sampling point. When the adjusted actual material intake volumetric flow rate is greater than the baseline volumetric flow rate and less than or equal to the rated volumetric conveying capacity, the difference between the adjusted actual material intake volumetric flow rate and the baseline volumetric flow rate is recorded as the compensation volumetric flow rate difference. When the adjusted actual material intake volumetric flow rate is less than or equal to the baseline volumetric flow rate, the compensation volumetric flow rate difference is recorded as zero. When the adjusted actual material intake volumetric flow rate is greater than the rated volumetric conveying capacity, the compensation volumetric flow rate difference is not recorded initially, and the reduction ratio adjustment process begins.

[0046] After obtaining the compensated volumetric flow rate difference for all path sampling points within the stable material intake path segment, the compensated volumetric flow rate difference for each path sampling point is multiplied by the preset movement time to obtain the compensated volume for that path sampling point. The compensated volumes for all path sampling points within the stable material intake path segment are summed to obtain the compensated flow rate increment. The compensated flow rate increment is compared with the flow rate deficit. When the compensated flow rate increment equals the flow rate deficit, the current translation amount, the current reduction ratio, the adjusted geometric cross-sectional area, and the adjusted rotational angular velocity are recorded.

[0047] When the compensation flow increment is less than the flow deficit, first increase the translation amount, and then reposition the bucket wheel cutting envelope according to the increased translation amount to reshape the adjusted closed cutting section profile. Recalculate the adjusted geometric cross-sectional area, the adjusted actual material handling volume flow rate, and the compensation flow increment. If the compensation flow increment is still less than the flow deficit after the translation amount reaches the preset upper limit, add a candidate path sampling point in the opposite direction of the preset rotary material handling path to the stable material handling path segment, and recalculate the compensation flow increment for the added stable material handling path segment. If the compensation flow increment is still less than the flow deficit after all candidate path sampling points are added to the stable material handling path segment, return to the baseline volume flow rate acquisition process, reduce the baseline volume flow rate, and recalculate the theoretical rotational angular velocity, velocity saturation segment, and flow deficit.

[0048] When the compensation flow increment exceeds the flow deficit, first reduce the translation amount, then reposition the bucket wheel cutting envelope according to the reduced translation amount, and recalculate the adjusted geometric cross-sectional area, adjusted actual material handling volume flow rate, and compensation flow increment. If, after reducing the translation amount to the initial translation amount, the compensation flow increment still exceeds the flow deficit, increase the reduction ratio, and recalculate the adjusted rotational velocity, adjusted actual material handling volume flow rate, and compensation flow increment according to the increased reduction ratio. If the adjusted actual material handling volume flow rate corresponding to any path sampling point exceeds the rated volume conveying capacity, increase the reduction ratio, and recalculate the adjusted rotational velocity and adjusted actual material handling volume flow rate. If, after increasing the reduction ratio, the adjusted actual material handling volume flow rate corresponding to any path sampling point is ≤ the baseline volume flow rate, decrease the reduction ratio or increase the translation amount and recalculate.

[0049] Following the sequence of first adjusting the translation amount, then adjusting the length of the stable material intake path segment, and finally adjusting the reduction ratio, the compensation flow increment is repeatedly calculated until the compensation flow increment equals the flow deficit, and the adjusted actual material intake volumetric flow rate at each sampling point within the stable material intake path segment is greater than the baseline volumetric flow rate and less than or equal to the rated volumetric conveying capacity. The translation amount, reduction ratio, adjusted geometric cross-sectional area, and adjusted rotational angular velocity at which the compensation flow increment equals the flow deficit are used as the bucket wheel cutting parameters corresponding to the stable material intake path segment.

[0050] The path sampling points before the stable material picking path section retain the theoretical rotational angular velocity; the path sampling points within the stable material picking path section use the adjusted rotational angular velocity; the path sampling points within the speed saturation section use Class II speed; and the path sampling points after the speed saturation section retain the theoretical rotational angular velocity. The results are arranged according to the replaced rotational angular velocities to form the adjusted rotational angular velocity command sequence.

[0051] This invention transforms the bucket wheel excavator's material handling path planning from simple geometric velocity planning to continuous material flow scheduling constrained by the mechanism's execution boundaries. When the bucket wheel excavator handles material at the edge of the stockpile or in thin material areas, the geometric cross-sectional area decreases rapidly. The theoretical rotational velocity calculated from the baseline volumetric flow rate may exceed the speed and acceleration range that the rotation mechanism can execute, resulting in local flow gaps during actual material handling. This gap is not caused by the stockpile geometry itself being unable to handle material, but rather by the difference between the theoretical speed command and the execution capability of the rotation mechanism. Therefore, during the path planning stage, inoperable sections are transformed into quantifiable flow gaps, and path sections that are still in a stable material handling state before the speed saturation section are used to undertake the compensation task. Stable material handling path sections have the conditions of continuous material contact and adjustable cutting cross-sections, which can form a compensating material flow in advance without exceeding the downstream conveying capacity, so that the flow reduction caused by the subsequent speed-limited section is offset by the pre-compensation amount. With this approach, path planning no longer uses the theoretical turning angular velocity directly as the execution command. Instead, it redistributes the material picking amount based on the continuity of the material flow over time. This reduces the impact of thin material areas at the edges on the downstream conveying flow and improves the consistency of path execution during unattended material picking.

[0052] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the present invention should still fall within the scope of the present invention.

Claims

1. A method for optimizing the material handling path of an unattended bucket wheel excavator based on industrial vision, characterized in that, Includes the following steps: The point cloud data of the material pile surface is collected by an industrial vision system, and a three-dimensional surface model of the material pile is established based on the point cloud data of the material pile surface. Obtain the baseline volumetric flow rate, the upper limit of the output speed of the rotary mechanism, and the upper limit of the acceleration; Calculate the geometric cross-sectional area point by point along the preset rotary material handling path, and calculate the theoretical rotational angular velocity based on the reference volumetric flow rate, geometric cross-sectional area and rotational radius; Based on the theoretical rotational angular velocity and the upper limit values ​​of velocity and acceleration, the velocity saturation section on the rotational material handling path is identified; The actual angular velocity curve is generated based on the upper limit values ​​of velocity and acceleration, and the flow deficit in the velocity saturation section is calculated by integration. Before the speed saturation zone, a stable material feeding path section is selected, the geometric cross-sectional area is increased and the rotational angular velocity is reduced, and the adjusted rotational angular velocity command sequence and bucket wheel cutting parameters are generated.

2. The method for optimizing the material handling path of an unattended bucket wheel excavator based on industrial vision according to claim 1, characterized in that, The process of establishing a three-dimensional surface model of the stockpile is as follows: Multiple two-dimensional images of the material pile are acquired using a variety of fixedly installed image acquisition devices; A stereo vision matching algorithm is used to process multi-view two-dimensional images to generate an initial dense point cloud. The initial dense point cloud is then subjected to statistical outlier removal filtering to obtain a filtered point cloud. The filtered point cloud is triangulated to generate a continuous surface mesh composed of triangular facets, which is then used as the three-dimensional surface model of the material pile.

3. The method for optimizing the material handling path of an unattended bucket wheel excavator based on industrial vision according to claim 1, characterized in that, The process of obtaining the baseline volumetric flow rate and the upper limits of the output speed and acceleration of the rotary mechanism is as follows: The reference material flow rate is received from the upper-level process control system of the bucket wheel excavator, the material bulk density is read from the material yard database, and the reference material flow rate is divided by the material bulk density to obtain the reference volumetric flow rate. Read the maximum output speed, maximum output torque, transmission ratio, and cantilever moment of inertia of the slewing motor from the parameters of the slewing mechanism drive device of the bucket wheel excavator; The upper limit of the angular velocity at the output end of the rotary mechanism is calculated based on the maximum output speed and transmission ratio of the rotary motor and used as the upper speed limit; The upper limit of the angular acceleration at the output end of the rotary mechanism is calculated based on the maximum output torque of the rotary motor, the transmission ratio, and the moment of inertia of the cantilever.

4. The method for optimizing the material handling path of an unattended bucket wheel excavator based on industrial vision according to claim 1, characterized in that, The process of calculating the theoretical rotational angular velocity is as follows: Transform the three-dimensional surface model of the material pile to the coordinate system of the bucket wheel excavator cantilever kinematics model; The preset rotary material handling path is discretized into multiple path sampling points, and the bucket wheel center coordinates, rotation radius and cutting section plane corresponding to each path sampling point are calculated. The three-dimensional surface model of the material pile is cut off by the cutting cross-section plane to obtain the broken line of the material pile cross-section; Generate the bucket wheel cutting envelope based on the bucket wheel center coordinates and the outer radius of the bucket teeth; The geometric cross-sectional area is obtained by combining the broken line of the material pile section with the cutting envelope of the bucket wheel, and the theoretical rotational angular velocity is calculated.

5. The method for optimizing the material handling path of an unattended bucket wheel excavator based on industrial vision according to claim 1, characterized in that, The process of identifying the velocity saturation zone is as follows: Arrange the theoretical rotational angular velocities corresponding to the sampling points of each path in the order of the preset rotary material handling path to obtain the angular velocity sequence; Calculate the absolute difference between adjacent theoretical angular velocities in the angular velocity sequence, and divide the absolute difference by the preset motion time to obtain the theoretical angular acceleration; Theoretical angular velocities with theoretical angular acceleration greater than the upper limit of acceleration are marked as abnormal velocities, and theoretical angular velocities with velocity greater than the upper limit of velocity are also marked as abnormal velocities. The path sampling points corresponding to continuous abnormal speeds form speed saturation sections.

6. The method for optimizing the material handling path of an unattended bucket wheel excavator based on industrial vision according to claim 5, characterized in that, The process of calculating the flow deficit is as follows: Sort the abnormal speeds according to the sampling points along the corresponding path; Each abnormal speed is replaced with the smaller value between the speed limit and the abnormal speed to obtain a class of speeds; Add the theoretical rotational angular velocity corresponding to the sampling point of the adjacent path before the starting point of the velocity saturation section to the first position of the sorting; adjust the first type of velocity by step according to the product of the upper limit of acceleration and the preset motion time to obtain the second type of velocity. Multiply the second type of velocity, the radius of gyration, and the geometric cross-sectional area to obtain the actual material intake volume flow rate; The flow deficit is calculated based on the difference between the baseline volumetric flow rate and the actual material intake volumetric flow rate.

7. The method for optimizing the material handling path of an unattended bucket wheel excavator based on industrial vision according to claim 1, characterized in that, The process of generating the adjusted angular velocity command sequence and bucket wheel cutting parameters is as follows: The rated volume conveying capacity of the downstream conveyor belt; Select continuous path sampling points that are not marked as speed saturation sections before the speed saturation section as stable material feeding path sections; The cutting section profile within the stable material handling path segment is translated to obtain the adjusted geometric cross-sectional area, and the corresponding rotational angular velocity is reduced. The compensation flow increment is calculated based on the adjusted geometric cross-sectional area, radius of gyration, and reduced angular velocity of gyration. The parameters when the compensation flow increment equals the flow deficit are used as the bucket wheel cutting parameters, and an adjusted slewing angular velocity command sequence is formed.

Images