Mine operation area vehicle cloud collaborative map updating method and system

Abstract

The invention discloses a mine operation area vehicle cloud collaborative map updating method and system, and the method comprises the steps: obtaining global pose scanning data of a mine operation area, and carrying out the semantic division of the global pose scanning data, and obtaining environment classification features; drawing a plurality of grid layers according to the environment classification features; constructing a cloud full-amount map by the grid map layer; controlling the working vehicle to execute the target task and obtain vehicle end scanning data, constructing a vehicle end increment map based on the vehicle end scanning data, performing semantic division on the vehicle end increment map to obtain local environment classification features, and updating each grid map layer in the cloud full-amount map according to the local environment classification features; and the mine operation area map is accurately updated in real time, so that the operation efficiency and safety of the unmanned working vehicle in the surface mine unloading area are ensured.

Description

Technical Field

[0001] The present invention belongs to the technical field of electronic map updating, and in particular relates to a vehicle-cloud collaborative map updating method and system for a mining operation area. Background Art

[0002] With the rapid development of industrial automation and intelligence, the field of open-pit mine transportation is undergoing a technological revolution. The application of unmanned driving technology aims to improve the safety, efficiency and economy of mine operations. In this context, many unmanned driving technology companies have launched their own unmanned transportation solutions for open-pit mines, promoting the development and application of related technologies.

[0003] However, when unmanned mining trucks operate in the unloading area of ​​open-pit mines, they usually face the situation of frequent changes in terrain caused by multiple vehicle unloading operations. First, in a multi-vehicle operation environment, vehicles will interfere with the update of maps. The existing technology fails to effectively remove or retain vehicle information for different operation tasks, affecting the accuracy of map updates. Secondly, the natural boundaries of the retaining walls and rock piles in the unloading area are irregular, and artificial boundaries will also be generated in the temporary roadblock area. It is difficult for the existing technology to identify natural boundaries and artificial boundaries and integrate them into the same map. In addition, due to the limited speed of bulldozers in bulldozing operations, the bulldozers cannot update the overall area boundaries and terrain in time, resulting in some unloading points being abandoned due to incomplete unloading, resulting in waste of unloading points, or some unloading points being over-unloaded, resulting in excessive material stacking. In summary, the existing technology cannot update the mine operation area map in a timely and accurate manner, which will affect the efficiency and safety of unmanned mining trucks in the unloading area of ​​open-pit mines. Summary of the invention

[0004] The present invention provides a vehicle-cloud collaborative map updating method and system for a mine operation area, which can update the mine operation area map in real time and accurately, thereby ensuring the efficiency and safety of unmanned mining trucks operating in the unloading area of ​​an open-pit mine.

[0005] In order to achieve the above object, the technical solution adopted by the present invention is:

[0006] A first aspect of the present invention provides a method for updating a vehicle-cloud collaborative map in a mine operation area, comprising:

[0007] Obtain global posture scanning data of the mine operation area, perform semantic segmentation on the global posture scanning data to obtain environmental classification features; draw a number of raster layers according to the environmental classification features; and construct a full cloud map from the raster layers;

[0008] Obtaining instance operation data of working vehicles in the mine operation area, and calculating similar feature values ​​based on the instance operation data and the preset vehicle model representation; classifying the working vehicles in the mine operation area according to the similar feature values ​​to obtain the working vehicle types, and adding the working vehicle operation attributes;

[0009] Use the quadtree algorithm to segment the working vehicles in the mine operation area to obtain vehicle distribution information; assign target tasks to the working vehicles based on vehicle distribution information, working vehicle type and operation attributes;

[0010] Control the working vehicle to perform the target task and obtain the vehicle-side scanning data, build the vehicle-side incremental map based on the vehicle-side scanning data, perform semantic division on the vehicle-side incremental map to obtain the local environment classification features, and update each raster layer in the full map on the cloud according to the local environment classification features.

[0011] Furthermore, semantic segmentation is performed on the global pose scan data to obtain environmental classification features. The process includes:

[0012] Establish a global coordinate system and construct a full point cloud map based on the global pose scanning data, perform ground segmentation and clustering on the pixels in the full point cloud map to obtain ground contour point sets and non-ground contour point sets;

[0013] The ground contour point set and the non-ground contour point set are input into the implicit function, and the expression formula is:

[0014]

[0015] In the formula, is an implicit function, k is the control coefficient; Indicates that the pixel sequence on the x-axis in the global coordinate system satisfies The p-degree polynomial basis function, ; is the first coordinate on the x-axis in the global coordinate system The position characteristics of pixels; Indicates that the pixel sequence on the x-axis in the global coordinate system satisfies The p-degree polynomial basis function, ; is the first coordinate on the y-axis in the global coordinate system The position characteristics of pixels; n is the number of pixels on the y-axis and x-axis in the global coordinate system;

[0016] The ground contour function and the non-ground contour function are obtained by solving the implicit function; based on the ground contour function and the non-ground contour function, the ground contour point set and the non-ground contour point set are semantically divided to obtain the environment classification features, which include boundary features, ground features, boundary features, mountain features and retaining wall features.

[0017] Furthermore, semantic segmentation of the ground contour point set and the non-ground contour point set is performed based on the ground contour function and the non-ground contour function to obtain the environment classification features. The process includes:

[0018] Obtain the implicit function value of each ground contour point from the ground contour function, and obtain the implicit function value of each non-ground contour point from the non-ground contour function;

[0019] When the difference between the implicit function value of the ground contour point and the implicit function value of the non-ground contour point is less than the set first threshold, the expression formula is: ; is the implicit function value difference of non-ground contour points; is the implicit function value of the ground contour point; is the first threshold; the ground contour points and the non-ground contour points are divided into boundary features;

[0020] Deleting the ground contour points as boundary features from the ground contour point set to obtain ground features;

[0021] The contour distribution data of the non-ground contour point set is calculated by the non-ground contour function; after deleting the non-ground contour points as boundary features from the non-ground contour point set, the non-ground contour point set is divided into mountain features and retaining wall features according to the contour distribution data.

[0022] Furthermore, the non-ground contour point set is divided into mountain features and retaining wall features according to the contour distribution data. The process includes:

[0023] The non-ground contour point set whose contour distribution data meets the regional division conditions is divided into regional environmental features; the contour distribution data includes contour gradient and contour altitude parameters; the regional division conditions include: the lowest altitude threshold of the contour line in the non-ground contour point concentration is greater than or equal to the highest altitude threshold of the contour line in the boundary feature, and the gradient of the contour line in the non-ground contour point concentration meets the monotonically decreasing trend;

[0024] If the gradient of the non-ground contour point set and the ground contour point set contour line merged meets the Lipschitz continuity condition, the regional environmental feature is classified as a retaining wall feature; otherwise, the regional environmental feature is classified as a mountain feature.

[0025] Furthermore, a number of raster layers are drawn according to the environmental classification characteristics, and the full cloud map is constructed from the raster layers. The process includes:

[0026] Draw the first grid layer according to the boundary features; read the manually set obstacle data and draw the second grid layer according to the obstacle data; draw the third grid layer according to the ground features, mountain features and retaining wall features;

[0027] A full cloud map is constructed based on the first raster layer, the second raster layer, and the third raster layer.

[0028] Furthermore, the process of establishing a vehicle model representation includes:

[0029] Obtain the coordinate position of the rear axle center A0 in the vehicle model, derive the position coordinates of the left front wheel center A1, the right front wheel center A2, the left rear wheel center A3 and the right rear wheel center A4 based on the coordinate position of the rear axle center A0, the body size and body posture of the working vehicle, and calculate the normal vector H1 of the plane passing through the left front wheel center A1, the right front wheel center A2, the left rear wheel center A3 and the right rear wheel center A4;

[0030] Based on the coordinate position of the rear axle center A0, the body size and body posture of the working vehicle, the position coordinates of the upper left front corner B1 and the upper right front corner B2 and the lower left rear corner B3 and the lower right rear corner B4 of the working vehicle are derived, and the normal vector H2 of the plane passing through the upper left front corner B1 and the upper right front corner B2 and the lower left rear corner B3 and the lower right rear corner B4 of the vehicle body is calculated;

[0031] Obtain the coordinate position of the front axle center C0 of the vehicle model, derive the coordinate positions of the left front wheel center C1, the right front wheel center C2, the left rear wheel center C3, and the right rear wheel center C4 based on the coordinate position of the front axle center C0, the body size and body posture of the working vehicle, and calculate the normal vector H3 of the plane passing through the left front wheel center C1, the right front wheel center C2, the left rear wheel center C3, and the right rear wheel center C4;

[0032] Based on the coordinate position of the front axle center C0, the body size and body posture of the working vehicle, the coordinate positions of the upper left corner D1, lower left corner D2, lower right corner D3, and upper right corner D4 of the working mechanism of the working vehicle are derived, and the normal vector H4 of the plane passing through the upper left corner D1, lower left corner D2, lower right corner D3, and upper right corner D4 of the working mechanism is calculated;

[0033] Normal vectors H1, H2, H3 and H4 are used as feature vectors of the vehicle model; the center position and speed of the vehicle model are obtained, and a vehicle model representation is constructed based on the center position and speed of the vehicle model and the feature vectors.

[0034] Further, similar feature values ​​are calculated based on the instance operation data and the preset vehicle model characterization, and the process includes:

[0035] The example operation data includes the center coordinates, vehicle speed, body coordinates and working mechanism coordinates of the working vehicle; the characteristic vector of the working vehicle is calculated according to the body coordinates and the working mechanism coordinates. , feature vector , feature vector and the feature vector ; According to the characteristic vector of the working vehicle , feature vector , feature vector , feature vector , center coordinates and vehicle speed to construct the working vehicle representation, denoted as ;In the formula, is the center position coordinate of the sth working vehicle at time t; is the speed of the sth working vehicle at time t; Characterize vehicles for work;

[0036] The vehicle model representation at time t is adjusted according to the center coordinates and vehicle speed of the working vehicle, and is recorded as ;In the formula, is the center position coordinate of the vth vehicle model at time t, is the speed of the vth vehicle model at time t; is the feature vector of the vehicle model; Characterize the vehicle model;

[0037] Based on the working vehicle characterization and vehicle model characterization at time t, similarity eigenvalues ​​are calculated and expressed as:

[0038]

[0039] In the formula, are similar eigenvalues; To set the weight coefficient.

[0040] Furthermore, the working attributes of the working vehicle include: unloading point location, working area attributes, vehicle fault code, motion attributes and target task attributes.

[0041] Furthermore, the quadtree algorithm is used to segment the working vehicles in the mine operation area to obtain vehicle distribution information. The process includes:

[0042] The center of the mine operation area is selected as the root node of the quadtree and the bounding box of the configuration space is set. The mine operation area is divided into smaller sub-areas according to the vehicle position data, which are recorded as quadtree nodes. The quadtree node splitting is repeated until the quadtree termination splitting condition is reached; the quadtree termination splitting condition includes: the number of working vehicles contained in each quadtree node is lower than a preset second threshold, and the split quadtree node reaches the bounding box of the set configuration space;

[0043] Assign each working vehicle to the quadtree node where it is located, count the number and type of vehicles in each quadtree node, and obtain the vehicle distribution information.

[0044] Furthermore, the vehicle-side incremental map is constructed based on the vehicle-side scanning data, the vehicle-side incremental map is semantically segmented to obtain local environment classification features, and each raster layer in the cloud-based full map is updated according to the local environment classification features. The process includes:

[0045] Based on the vehicle-side scanning data, a vehicle-side incremental map is constructed, local features of obstacles are extracted from the vehicle-side incremental map, and semantic segmentation is performed on the vehicle-side incremental map to obtain local features of boundaries, local features of the ground, local features of mountains, and local features of retaining walls; local environment classification features are composed of local features of obstacles, local features of boundaries, local features of the ground, local features of mountains, and local features of retaining walls;

[0046] When the map update conditions are met, the first grid layer is updated according to the local features of the boundary, the second grid layer is updated according to the local features of the obstacle, and the third grid layer is updated according to the local features of the ground, the local features of the mountain, and the local features of the retaining wall;

[0047] The map update conditions include that the number of unloading times at the unloading point is greater than a set third threshold, the unloading of the working vehicle is abnormal, or the data before the specified time is discarded;

[0048] The full cloud map is redrawn based on the updated first raster layer, second raster layer, and third raster layer.

[0049] A second aspect of the present invention provides a vehicle-cloud collaborative map updating system for a mine operation area, which is used to execute the vehicle-cloud collaborative map updating method for a mine operation area described in the first aspect. The vehicle-cloud collaborative map updating system for a mine operation area comprises:

[0050] The cloud-based full-volume map construction module is used to obtain the global posture scanning data of the mine operation area, perform semantic segmentation on the global posture scanning data to obtain environmental classification features; draw a number of raster layers according to the environmental classification features; and construct the cloud-based full-volume map from the raster layers;

[0051] The vehicle-side data acquisition module is used to obtain the instance operation data of the working vehicles in the mine operation area, calculate the similarity feature value based on the instance operation data and the preset vehicle model representation; classify the working vehicles in the mine operation area according to the similarity feature value to obtain the working vehicle type, and add the working vehicle's operation attributes;

[0052] The task allocation module uses the quadtree algorithm to segment the working vehicles in the mine operation area to obtain vehicle distribution information; it allocates target tasks to the working vehicles based on the vehicle distribution information, working vehicle type and operation attributes;

[0053] The map update module is used to control the working vehicle to perform the target task and obtain the vehicle-side scanning data, build the vehicle-side incremental map based on the vehicle-side scanning data, perform semantic division on the vehicle-side incremental map to obtain local environment classification features, and update each raster layer in the full map on the cloud according to the local environment classification features.

[0054] The third aspect of the present invention provides an electronic device, including a storage medium and a processor; the storage medium is used to store instructions; the processor is used to operate according to the instructions to execute the vehicle-cloud collaborative map update method for the mining operation area described in the first aspect.

[0055] Compared with the prior art, the present invention has the following beneficial effects:

[0056] The present invention controls the working vehicle to perform the target task and obtains the vehicle-side scanning data, constructs the vehicle-side incremental map based on the vehicle-side scanning data, performs semantic division on the vehicle-side incremental map to obtain the local environment classification characteristics, and updates each raster layer in the cloud-based full map according to the local environment classification characteristics; the real-time updated vehicle and environmental information helps to avoid vehicle and equipment conflicts during the operation process, improves operation safety, enhances the ability to respond quickly to sudden environmental changes, and can adjust the operation plan in time to avoid potential safety accidents.

[0057] The present invention obtains instance operation data of working vehicles in a mine operation area, calculates similar feature values ​​according to the instance operation data and a preset vehicle model characterization; classifies the working vehicles in the mine operation area according to the similar feature values, and adds the working attributes of the working vehicles; uses a quadtree algorithm to perform regional segmentation on the working vehicles in the mine operation area to obtain vehicle distribution information; assigns target tasks to the working vehicles according to the vehicle distribution information, the type of working vehicles and the working attributes; more optimized path planning can be performed through accurate vehicle and environmental information, and the workload of each working vehicle can be balanced and resource utilization can be improved through reasonable task allocation and vehicle scheduling.

[0058] The present invention obtains global posture scanning data of a mine operation area, performs semantic division on the global posture scanning data to obtain environmental classification features; draws a plurality of grid layers according to the environmental classification features; constructs a cloud-based full map from the grid layers; updates each grid layer in the cloud-based full map according to the local environmental classification features; can adapt to operation areas of different sizes, has good scalability, can flexibly respond to changes in operation requirements and environment, and quickly adjust operation plans. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 A flowchart of a method for updating a vehicle-cloud collaborative map in a mine operation area provided in Example 1 of the present invention;

[0060] Figure 2This is a contour distribution map of non-ground contours and ground contours provided in Example 1 of the present invention.

[0061] Figure 3 A flowchart of drawing a full cloud map provided in Example 1 of the present invention;

[0062] Figure 4 A structural diagram of building a vehicle model provided in Example 1 of the present invention;

[0063] Figure 5 A structural diagram of a full cloud map provided in Example 1 of the present invention;

[0064] Figure 6 A flow chart of a vehicle-cloud collaborative map updating system for a mine operation area provided in Example 2 of the present invention;

[0065] Figure 7 This is a structural diagram of the work vehicle management module provided in Example 2 of the present invention. DETAILED DESCRIPTION

[0066] The present invention will be further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and cannot be used to limit the protection scope of the present invention.

[0067] Example 1

[0068] This implementation provides a method for updating a mine operation area vehicle-cloud collaborative map, including:

[0069] Obtain the global posture scanning data of the mine operation area, and perform semantic segmentation on the global posture scanning data to obtain environmental classification features. The process includes:

[0070] Establish a global coordinate system and construct a full point cloud map based on the global pose scanning data, perform ground segmentation and clustering on the pixels in the full point cloud map to obtain ground contour point sets and non-ground contour point sets;

[0071] The ground contour point set and the non-ground contour point set are input into the implicit function, and the expression formula is:

[0072]

[0073] In the formula, is an implicit function, k is the control coefficient; Indicates that the pixel sequence on the x-axis in the global coordinate system satisfies The p-degree polynomial basis function, ; is the first coordinate on the x-axis in the global coordinate system The position characteristics of pixels; Indicates that the pixel sequence on the x-axis in the global coordinate system satisfies The p-degree polynomial basis function, ; is the first coordinate on the y-axis in the global coordinate system The position characteristics of pixels; n is the number of pixels on the y-axis and x-axis in the global coordinate system;

[0074] The ground contour function and the non-ground contour function are obtained by solving the implicit function; the ground contour point set and the non-ground contour point set are semantically divided based on the ground contour function and the non-ground contour function to obtain the environmental classification features. The process includes:

[0075] The environmental classification features include boundary features, ground features, boundary features, mountain features and retaining wall features;

[0076] Obtain the implicit function value of each ground contour point from the ground contour function, and obtain the implicit function value of each non-ground contour point from the non-ground contour function;

[0077] When the difference between the implicit function value of the ground contour point and the implicit function value of the non-ground contour point is less than the set first threshold, the expression formula is: ; is the implicit function value difference of non-ground contour points; is the implicit function value of the ground contour point; is the first threshold; the ground contour points and the non-ground contour points are divided into boundary features;

[0078] Deleting the ground contour points as boundary features from the ground contour point set to obtain ground features;

[0079] The contour distribution data of the non-ground contour point set is calculated by the non-ground contour function; after deleting the non-ground contour points as boundary features from the non-ground contour point set, the non-ground contour point set is divided into mountain features and retaining wall features according to the contour distribution data. The process includes:

[0080] The non-ground contour point set whose contour distribution data meets the regional division conditions is divided into regional environmental features; the contour distribution data includes contour gradient and contour altitude parameters; the regional division conditions include: the lowest altitude threshold of the contour line in the non-ground contour point concentration is greater than or equal to the highest altitude threshold of the contour line in the boundary feature, and the gradient of the contour line in the non-ground contour point concentration meets the monotonically decreasing trend;

[0081] If the gradient of the non-ground contour point set and the ground contour point set contour line merged meets the Lipschitz continuity condition, the regional environmental feature is classified as a retaining wall feature; otherwise, the regional environmental feature is classified as a mountain feature.

[0082] Draw several raster layers according to the environmental classification characteristics, and construct a full cloud map from the raster layers. The process includes:

[0083] Draw a first grid layer according to the boundary features; read artificially set obstacle data, the obstacle data includes but is not limited to faulty vehicles and large rocks, and draw a second grid layer according to the obstacle data; draw a third grid layer according to ground features, mountain features and retaining wall features;

[0084] A full cloud map is constructed based on the first raster layer, the second raster layer and the third raster layer; the business content of the full cloud map includes: unloading waiting area, retaining wall area, mountain area, controlled area (such as vehicle breakdown); entry path, exit path; unloading point; area boundary.

[0085] The types of working vehicles include but are not limited to mining trucks and bulldozers. The process of establishing a vehicle model representation based on the types of working vehicles includes:

[0086] Obtain the coordinate position of the rear axle center A0 in the vehicle model, derive the position coordinates of the left front wheel center A1, the right front wheel center A2, the left rear wheel center A3 and the right rear wheel center A4 based on the coordinate position of the rear axle center A0, the body size and body posture of the working vehicle, and calculate the normal vector H1 of the plane passing through the left front wheel center A1, the right front wheel center A2, the left rear wheel center A3 and the right rear wheel center A4;

[0087] Based on the coordinate position of the rear axle center A0, the body size and body posture of the working vehicle, the position coordinates of the upper left front corner B1 and the upper right front corner B2 and the lower left rear corner B3 and the lower right rear corner B4 of the working vehicle are derived, and the normal vector H2 of the plane passing through the upper left front corner B1 and the upper right front corner B2 and the lower left rear corner B3 and the lower right rear corner B4 of the vehicle body is calculated;

[0088] Obtain the coordinate position of the front axle center C0 of the vehicle model, derive the coordinate positions of the left front wheel center C1, the right front wheel center C2, the left rear wheel center C3, and the right rear wheel center C4 based on the coordinate position of the front axle center C0, the body size and body posture of the working vehicle, and calculate the normal vector H3 of the plane passing through the left front wheel center C1, the right front wheel center C2, the left rear wheel center C3, and the right rear wheel center C4;

[0089] Based on the coordinate position of the front axle center C0, the body size and body posture of the working vehicle, the coordinate positions of the upper left corner D1, lower left corner D2, lower right corner D3, and upper right corner D4 of the working mechanism of the working vehicle are derived, and the normal vector H4 of the plane passing through the upper left corner D1, lower left corner D2, lower right corner D3, and upper right corner D4 of the working mechanism is calculated;

[0090] Normal vectors H1, H2, H3 and H4 are used as feature vectors of the vehicle model; the center position and speed of the vehicle model are obtained, and a vehicle model representation is constructed based on the center position and speed of the vehicle model and the feature vectors.

[0091] Obtain the instance operation data of the working vehicles in the mine operation area, and calculate the similarity characteristic value based on the instance operation data and the preset vehicle model representation. The process includes:

[0092] The example operation data includes the center coordinates, vehicle speed, body coordinates and working mechanism coordinates of the working vehicle; the characteristic vector of the working vehicle is calculated according to the body coordinates and the working mechanism coordinates. , feature vector , feature vector and the feature vector ; According to the characteristic vector of the working vehicle , feature vector , feature vector , feature vector , center coordinates and vehicle speed to construct the working vehicle representation, denoted as ;In the formula, is the center position coordinate of the sth working vehicle at time t; is the speed of the sth working vehicle at time t; Characterize vehicles for work;

[0093] The vehicle model representation at time t is adjusted according to the center coordinates and vehicle speed of the working vehicle, and is recorded as ;In the formula, is the center position coordinate of the vth vehicle model at time t, is the speed of the vth vehicle model at time t; is the feature vector of the vehicle model; Characterize the vehicle model;

[0094] Based on the working vehicle characterization and vehicle model characterization at time t, similarity eigenvalues ​​are calculated and expressed as:

[0095]

[0096] In the formula, are similar eigenvalues; To set the weight coefficient.

[0097] According to similar feature values, the working vehicles in the mine operation area are classified to obtain the working vehicle type (mine truck value is 1 / bulldozer value is 2 / excavator value is 3), and the working vehicle operation attributes are added; the working vehicle operation attributes include: operating area attribute [Area_ID], motion attribute [Motion_Attribute], vehicle fault code [Vehicle_Code] (fault 1 / normal-1), [Operation_Attribute] (waiting for unloading value is 1 / unloading value is 2 / unloading completed value is 3), task target attribute [Unknow_Attribute] (known target is 255, unknown target is 0) and unloading point location [Unload_P_X, Unload_P_Y, Unload_P_Z].

[0098] The quadtree algorithm is used to segment the working vehicles in the mine operation area to obtain vehicle distribution information, and target tasks are assigned to the working vehicles according to the vehicle distribution information. The process includes:

[0099] The center of the mine operation area is selected as the root node of the quadtree and the bounding box of the configuration space is set. The mine operation area is divided into smaller sub-areas according to the vehicle position data, which are recorded as quadtree nodes. The quadtree node splitting is repeated until the quadtree termination splitting condition is reached; the quadtree termination splitting condition includes: the number of working vehicles contained in each quadtree node is lower than a preset second threshold, and the split quadtree node reaches the bounding box of the set configuration space;

[0100] Assign each working vehicle to the quadtree node where it is located, count the number and type of vehicles in each quadtree node, and obtain the vehicle distribution information; assign target tasks to working vehicles based on vehicle distribution information, working vehicle type, and job attributes; more optimized path planning can be performed through accurate vehicle and environmental information, and reasonable task allocation and vehicle scheduling can balance the workload of each working vehicle and improve resource utilization.

[0101] Control the working vehicle to perform the target task and obtain the vehicle-side scanning data, build the vehicle-side incremental map based on the vehicle-side scanning data, perform semantic segmentation on the vehicle-side incremental map to obtain the local environment classification features, and update each raster layer in the cloud-based full map according to the local environment classification features. The process includes:

[0102] Based on the vehicle-side scanning data, a vehicle-side incremental map is constructed, local features of obstacles are extracted from the vehicle-side incremental map, and semantic segmentation is performed on the vehicle-side incremental map to obtain local features of boundaries, local features of the ground, local features of mountains, and local features of retaining walls; local environment classification features are composed of local features of obstacles, local features of boundaries, local features of the ground, local features of mountains, and local features of retaining walls;

[0103] When the map update conditions are met, the first grid layer is updated according to the local features of the boundary, the second grid layer is updated according to the local features of the obstacle, and the third grid layer is updated according to the local features of the ground, the local features of the mountain, and the local features of the retaining wall;

[0104] The map update conditions include that the number of unloading times at the unloading point is greater than a set third threshold, the unloading of the working vehicle is abnormal, or the data before the specified time is discarded;

[0105] The full cloud map is redrawn based on the updated first raster layer, second raster layer and third raster layer; it can adapt to work areas of different sizes, has good scalability, can flexibly respond to changes in work needs and environment, and quickly adjust work plans.

[0106] The vehicle and environmental information that can be updated in real time in this embodiment helps to avoid vehicle and equipment conflicts during operations, improves operation safety, enhances the ability to respond quickly to sudden environmental changes, and can adjust operation plans in a timely manner to avoid potential safety accidents.

[0107] Example 2

[0108] This embodiment provides a mine operation area vehicle-cloud collaborative map updating system, which is used to execute the mine operation area vehicle-cloud collaborative map updating method described in Embodiment 1. The mine operation area vehicle-cloud collaborative map updating system includes:

[0109] The cloud-based full-volume map construction module is used to obtain the global posture scanning data of the mine operation area, perform semantic segmentation on the global posture scanning data to obtain environmental classification features; draw a number of raster layers according to the environmental classification features; and construct the cloud-based full-volume map from the raster layers;

[0110] The vehicle-side data acquisition module is used to obtain the instance operation data of the working vehicles in the mine operation area, calculate the similarity feature value based on the instance operation data and the preset vehicle model representation; classify the working vehicles in the mine operation area according to the similarity feature value, and add the working attributes of the working vehicles;

[0111] A working vehicle management module is used to judge the fault state, motion state, unloading state, and target attributes of the working vehicle; the working vehicle management module is provided with a semantic category index, an instance object index, intermediate business data, a rule matrix, and result business data; the environment classification features obtained by semantic division are indexed to construct a semantic category index, and the semantic category index is used to represent the name of the environment classification feature, including but not limited to boundary features, ground features, boundary features, mountain features, and retaining wall features; the instance object index is an index value randomly constructed for the instance vehicle, and is used to represent the name of the current instance vehicle; the intermediate business data includes the working vehicle representation and operation attributes of the current working vehicle; the rule matrix judges the fault state, motion state, unloading state, and target attributes of the working vehicle according to the intermediate business data, and selects to delete, retain, and assign attributes to the intermediate business data of the working vehicle according to the judgment result, and stores it in the result business data;

[0112] The task allocation module uses the quadtree algorithm to segment the working vehicles in the mine operation area to obtain vehicle distribution information; it allocates target tasks to the working vehicles based on the vehicle distribution information, working vehicle type and operation attributes;

[0113] The map update module is used to control the working vehicle to perform the target task and obtain the vehicle-side scanning data, build the vehicle-side incremental map based on the vehicle-side scanning data, perform semantic division on the vehicle-side incremental map to obtain local environment classification features, and update each raster layer in the full map on the cloud according to the local environment classification features.

[0114] This embodiment can update the mine operation area map in real time and accurately, thereby ensuring the efficiency and safety of unmanned mining trucks operating in the unloading area of ​​an open-pit mine.

[0115] Example 3

[0116] This embodiment provides an electronic device, including a storage medium and a processor; the storage medium is used to store instructions; the processor is used to operate according to the instructions to execute the mine operation area vehicle-cloud collaborative map update method described in Example 1.

[0117] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0118] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0119] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0120] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0121] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A vehicle-cloud collaborative map updating method for a mine operation area, characterized in that: include: Obtain global posture scanning data of the mine operation area, and perform semantic segmentation on the global posture scanning data to obtain environmental classification features; Draw a number of raster layers according to the environmental classification characteristics; construct a full cloud map from the raster layers; Obtaining instance operation data of working vehicles in the mine operation area, and calculating similarity feature values ​​based on the instance operation data and the preset vehicle model representation; Classify the working vehicles in the mine operation area according to similar feature values ​​to obtain the working vehicle types, and add the working vehicle operation attributes; Use the quadtree algorithm to segment the working vehicles in the mine operation area to obtain vehicle distribution information; assign target tasks to the working vehicles based on vehicle distribution information, working vehicle type and operation attributes; Control the working vehicle to perform the target task and obtain the vehicle-side scanning data, build the vehicle-side incremental map based on the vehicle-side scanning data, perform semantic division on the vehicle-side incremental map to obtain the local environment classification features, and update each raster layer in the full map on the cloud according to the local environment classification features.

2. The method for updating the mine operation area vehicle-cloud collaborative map according to claim 1 is characterized in that: The global pose scan data is semantically segmented to obtain environmental classification features. The process includes: Establish a global coordinate system and construct a full point cloud map based on the global pose scanning data, perform ground segmentation and clustering on the pixels in the full point cloud map to obtain ground contour point sets and non-ground contour point sets; The ground contour point set and the non-ground contour point set are input into the implicit function, and the expression formula is: ； In the formula, is an implicit function, k is the control coefficient; Indicates that the pixel sequence on the x-axis in the global coordinate system satisfies The p-degree polynomial basis function, ; is the x-axis in the global coordinate system The position characteristics of pixels; Indicates that the pixel sequence on the x-axis in the global coordinate system satisfies The p-degree polynomial basis function, ; is the first coordinate on the y-axis in the global coordinate system The position characteristics of pixels; n is the number of pixels on the y-axis and x-axis in the global coordinate system; The ground contour function and the non-ground contour function are obtained by solving the implicit function; based on the ground contour function and the non-ground contour function, the ground contour point set and the non-ground contour point set are semantically divided to obtain the environment classification features, which include boundary features, ground features, boundary features, mountain features and retaining wall features.

3. The method for updating the mine operation area vehicle-cloud collaborative map according to claim 2 is characterized in that: Based on the ground contour function and the non-ground contour function, the ground contour point set and the non-ground contour point set are semantically divided to obtain the environmental classification features. The process includes: Obtain the implicit function value of each ground contour point from the ground contour function, and obtain the implicit function value of each non-ground contour point from the non-ground contour function; When the difference between the implicit function value of the ground contour point and the implicit function value of the non-ground contour point is less than the set first threshold, the expression formula is: ; is the implicit function value difference of non-ground contour points; is the implicit function value of the ground contour point; is the first threshold; the ground contour points and the non-ground contour points are divided into boundary features; Deleting the ground contour points as boundary features from the ground contour point set to obtain ground features; The contour distribution data of the non-ground contour point set is calculated by the non-ground contour function; after deleting the non-ground contour points as boundary features from the non-ground contour point set, the non-ground contour point set is divided into mountain features and retaining wall features according to the contour distribution data.

4. The method for updating a mine operation area vehicle-cloud collaborative map according to claim 3 is characterized in that: The non-ground contour point set is divided into mountain features and retaining wall features according to the contour distribution data. The process includes: The non-ground contour point set whose contour distribution data meets the regional division conditions is divided into regional environmental features; the contour distribution data includes contour gradient and contour altitude parameters; the regional division conditions include: the lowest altitude threshold of the contour line in the non-ground contour point concentration is greater than or equal to the highest altitude threshold of the contour line in the boundary feature, and the gradient of the contour line in the non-ground contour point concentration meets the monotonically decreasing trend; If the gradient of the non-ground contour point set and the ground contour point set contour line merged meets the Lipschitz continuity condition, the regional environmental feature is classified as a retaining wall feature; otherwise, the regional environmental feature is classified as a mountain feature.

5. The method for updating the mine operation area vehicle-cloud collaborative map according to claim 2 is characterized in that: Draw several raster layers according to the environmental classification characteristics, and construct a full cloud map from the raster layers. The process includes: Draw the first grid layer according to the boundary features; read the manually set obstacle data and draw the second grid layer according to the obstacle data; draw the third grid layer according to the ground features, mountain features and retaining wall features; A full cloud map is constructed based on the first raster layer, the second raster layer, and the third raster layer.

6. The method for updating the mine operation area vehicle-cloud collaborative map according to claim 1 is characterized in that: The process of building a vehicle model representation includes: Obtain the coordinate position of the rear axle center A0 in the vehicle model, derive the position coordinates of the left front wheel center A1, the right front wheel center A2, the left rear wheel center A3 and the right rear wheel center A4 based on the coordinate position of the rear axle center A0, the body size and body posture of the working vehicle, and calculate the normal vector H1 of the plane passing through the left front wheel center A1, the right front wheel center A2, the left rear wheel center A3 and the right rear wheel center A4; Based on the coordinate position of the rear axle center A0, the body size and body posture of the working vehicle, the position coordinates of the upper left front corner B1 and the upper right front corner B2 and the lower left rear corner B3 and the lower right rear corner B4 of the working vehicle are derived, and the normal vector H2 of the plane passing through the upper left front corner B1 and the upper right front corner B2 and the lower left rear corner B3 and the lower right rear corner B4 of the vehicle body is calculated; Obtain the coordinate position of the front axle center C0 of the vehicle model, derive the coordinate positions of the left front wheel center C1, the right front wheel center C2, the left rear wheel center C3, and the right rear wheel center C4 based on the coordinate position of the front axle center C0, the body size and body posture of the working vehicle, and calculate the normal vector H3 of the plane passing through the left front wheel center C1, the right front wheel center C2, the left rear wheel center C3, and the right rear wheel center C4; Based on the coordinate position of the front axle center C0, the body size and body posture of the working vehicle, the coordinate positions of the upper left corner D1, lower left corner D2, lower right corner D3, and upper right corner D4 of the working mechanism of the working vehicle are derived, and the normal vector H4 of the plane passing through the upper left corner D1, lower left corner D2, lower right corner D3, and upper right corner D4 of the working mechanism is calculated; Normal vectors H1, H2, H3 and H4 are used as feature vectors of the vehicle model; the center position and speed of the vehicle model are obtained, and a vehicle model representation is constructed based on the center position and speed of the vehicle model and the feature vectors.

7. The method for updating a mine operation area vehicle-cloud collaborative map according to claim 6, characterized in that: The similarity characteristic values ​​are calculated based on the instance operation data and the preset vehicle model characterization. The process includes: The example operation data includes the center coordinates, vehicle speed, body coordinates and working mechanism coordinates of the working vehicle; the characteristic vector of the working vehicle is calculated according to the body coordinates and the working mechanism coordinates. , feature vector , feature vector and the feature vector ; According to the characteristic vector of the working vehicle , feature vector , feature vector , feature vector , center coordinates and vehicle speed to construct the working vehicle representation, denoted as ;In the formula, is the center position coordinate of the sth working vehicle at time t; is the speed of the sth working vehicle at time t; Characterize vehicles for work; The vehicle model representation at time t is adjusted according to the center coordinates and vehicle speed of the working vehicle, and is recorded as ;In the formula, is the center position coordinate of the vth vehicle model at time t, is the speed of the vth vehicle model at time t; is the feature vector of the vehicle model; Characterize the vehicle model; Based on the working vehicle characterization and vehicle model characterization at time t, similarity eigenvalues ​​are calculated and expressed as: ； In the formula, are similar eigenvalues; To set the weight coefficient.

8. The method for updating a mine operation area vehicle-cloud collaborative map according to claim 1, characterized in that: The working attributes of the working vehicle include: unloading point location, working area attributes, vehicle fault code, motion attributes and target task attributes.

9. The method for updating a mine operation area vehicle-cloud collaborative map according to claim 1, characterized in that: The quadtree algorithm is used to segment the working vehicles in the mine operation area to obtain vehicle distribution information. The process includes: The center of the mine operation area is selected as the root node of the quadtree and the bounding box of the configuration space is set. The mine operation area is divided into smaller sub-areas according to the vehicle position data, which are recorded as quadtree nodes. The quadtree node splitting is repeated until the quadtree termination splitting condition is reached; the quadtree termination splitting condition includes: the number of working vehicles contained in each quadtree node is lower than a preset second threshold value, and the split quadtree node reaches the bounding box of the set configuration space; Assign each working vehicle to the quadtree node where it is located, count the number and type of vehicles in each quadtree node, and obtain the vehicle distribution information.

10. The method for updating a mine operation area vehicle-cloud collaborative map according to claim 1, characterized in that: The vehicle-side incremental map is constructed based on the vehicle-side scanning data, and the vehicle-side incremental map is semantically divided to obtain the local environment classification features. The raster layers in the cloud-based full map are updated according to the local environment classification features. The process includes: Based on the vehicle-side scanning data, a vehicle-side incremental map is constructed, local features of obstacles are extracted from the vehicle-side incremental map, and semantic segmentation is performed on the vehicle-side incremental map to obtain local features of boundaries, local features of the ground, local features of mountains, and local features of retaining walls; local environment classification features are composed of local features of obstacles, local features of boundaries, local features of the ground, local features of mountains, and local features of retaining walls; When the map update conditions are met, the first grid layer is updated according to the local features of the boundary, the second grid layer is updated according to the local features of the obstacle, and the third grid layer is updated according to the local features of the ground, the local features of the mountain, and the local features of the retaining wall; The map update conditions include that the number of unloading times at the unloading point is greater than a set third threshold, the unloading of the working vehicle is abnormal, or the data before the specified time is discarded; The full cloud map is redrawn based on the updated first raster layer, second raster layer, and third raster layer.

11. A vehicle-cloud collaborative map updating system for a mining operation area, characterized in that: include: The cloud-based full-map construction module is used to obtain the global posture scanning data of the mining operation area and perform semantic segmentation on the global posture scanning data to obtain environmental classification features; Draw a number of raster layers according to the environmental classification characteristics; construct a full cloud map from the raster layers; The vehicle-side data acquisition module is used to obtain the instance operation data of the working vehicles in the mine operation area, and calculate the similarity feature value based on the instance operation data and the preset vehicle model representation; Classify the working vehicles in the mine operation area according to similar feature values ​​to obtain the working vehicle types, and add the working vehicle operation attributes; The task allocation module uses the quadtree algorithm to segment the working vehicles in the mine operation area to obtain vehicle distribution information; it allocates target tasks to the working vehicles based on the vehicle distribution information, working vehicle type and operation attributes; The map update module is used to control the working vehicle to perform the target task and obtain the vehicle-side scanning data, build the vehicle-side incremental map based on the vehicle-side scanning data, perform semantic division on the vehicle-side incremental map to obtain local environment classification features, and update each raster layer in the full map on the cloud according to the local environment classification features.

12. An electronic device comprising a storage medium and a processor; the storage medium is used to store instructions; the processor is used to operate according to the instructions to execute the vehicle-cloud collaborative map update method for a mining operation area as described in any one of claims 1 to 10.

Images