Obstacle collision detection method, device and unmanned vehicle

Abstract

The present disclosure provides an obstacle collision detection method, device and unmanned vehicle, comprising: traversing the trajectory points on the driving trajectory in front of the unmanned vehicle, and for each trajectory point, determining the normal direction of the driving trajectory at the trajectory point, and the first detection range corresponding to the trajectory point along the normal direction; searching the obstacle data in the storage space of the obstacle data according to the first detection range corresponding to each trajectory point, and determining whether there is spatial overlap between the obstacle data and any first detection range; in the case of spatial overlap, determining whether the unmanned vehicle and the obstacle will collide based on the position of the spatial overlap. The present disclosure can detect whether there is an actual geometric collision between the driving lane of the unmanned vehicle on the driving trajectory and an irregular obstacle. Accurate and effective collision detection is performed on obstacles of various shapes, effectively utilizing the passage space, and avoiding vehicle mis-stopping due to erroneous collision detection results.

Description

Technical Field

[0001] The present disclosure relates to the field of unmanned driving technology, and in particular to an obstacle collision detection method, device, and unmanned vehicle. Background Art

[0002] The operational scenarios for unmanned driving in mines are relatively complex, often featuring a variety of irregularly shaped obstacles, such as large fallen rocks, soil piles, and retaining walls. Traditionally, sphere envelopes, AABBbox envelopes, OBBbox envelopes, and grid map construction methods are used to detect collisions with obstacles. However, the sphere envelope method can cause areas where no collisions actually occur to be mistakenly detected as having occurred, preventing vehicles from passing. The AABBbox envelope method creates large, non-drivable areas in narrow mine scenarios, resulting in wasted space. The OBBbox envelope method cannot produce effective collision detection results for irregular concave polygonal obstacles. The grid map construction method is computationally intensive and lacks tracking properties, making it inconvenient for use with other modules.

[0003] Based on the above problems, there is an urgent need to provide an obstacle collision detection method to accurately and effectively detect obstacles of various shapes. Summary of the Invention

[0004] The embodiments of the present disclosure provide an obstacle collision detection method, device, and unmanned vehicle to solve the existing problem of being unable to effectively detect collisions with irregularly shaped obstacles.

[0005] In view of the above problems, in a first aspect, an embodiment of the present disclosure provides a method for detecting an obstacle collision, comprising:

[0006] The unmanned vehicle detects obstacles while driving and stores the detected obstacle data according to a preset data structure;

[0007] Traversing the trajectory points on the driving trajectory in front of the unmanned vehicle, and for each trajectory point, determining the normal direction of the driving trajectory at the trajectory point and a first detection range corresponding to the trajectory point along the normal direction;

[0008] Searching the obstacle data in the obstacle data storage space according to the first detection ranges corresponding to the respective trajectory points to determine whether any obstacle data spatially overlaps with any of the first detection ranges;

[0009] In the case of spatial overlap, whether the unmanned vehicle will collide with the obstacle is determined based on the position of the spatial overlap.

[0010] In conjunction with the first aspect, in one possible implementation, determining, for each trajectory point, the normal direction of the driving trajectory at the trajectory point and the first detection range corresponding to the trajectory point along the normal direction includes:

[0011] For each trajectory point, determining the normal direction of the driving trajectory at the trajectory point and a first detection line segment of the trajectory point along the normal direction; wherein the length of the first detection line segment is related to a preset detection boundary;

[0012] The step of searching the obstacle data in the obstacle data storage space according to the first detection ranges corresponding to the respective trajectory points to determine whether the obstacle data spatially overlaps with any of the first detection ranges includes:

[0013] Discretely process the first detection line segments corresponding to the respective trajectory points to determine a plurality of first detection points located on the respective first detection line segments;

[0014] For each first detection point, searching the obstacle data in a storage space of the obstacle data based on the hierarchical spatial index structure of the obstacle data to determine whether there is spatial overlap between the first detection point and the obstacle data;

[0015] The hierarchical spatial index structure of the obstacle data is related to the preset data structure based on which the obstacles are stored.

[0016] In conjunction with the first aspect, in one possible implementation, for each first detection point, searching the obstacle data in the storage space of the obstacle data based on the hierarchical spatial index structure of the obstacle data to determine whether there is spatial overlap between the first detection point and the obstacle data includes:

[0017] Traversing the first detection points of each first detection line segment, and determining, for each first detection point, the regional position coordinates of the area covered by the first detection point;

[0018] Searching the storage space of the obstacle data for the position coordinates of the obstacles contained in the obstacle data using a range search method according to the hierarchical spatial index structure of the obstacle data to determine whether any obstacle data is within the regional position coordinate range;

[0019] If so, it is determined that the first detection point has spatial overlap with the obstacle data.

[0020] In conjunction with the first aspect, in one possible implementation, for each first detection point, searching the obstacle data in the storage space of the obstacle data based on the hierarchical spatial index structure of the obstacle data to determine whether there is spatial overlap between the first detection point and the obstacle data includes:

[0021] Traversing the first detection points of each first detection line segment, and for each first detection point, searching the storage space of the obstacle data for the position coordinates of the obstacle contained in the obstacle data using a nearest neighbor search method based on the hierarchical spatial index structure of the obstacle data;

[0022] When a first distance between the position coordinates of the first detection point and the position coordinates of any obstacle is less than a first threshold, it is determined that there is spatial overlap between the first detection point and the obstacle data.

[0023] In conjunction with the first aspect, in one possible implementation, when there is spatial overlap, determining whether the unmanned vehicle will collide with the obstacle based on the position of the spatial overlap includes:

[0024] In a case where the first detection point spatially overlaps with the obstacle data, determining a second detection point among the first detection points that spatially overlaps with the obstacle data;

[0025] Whether the unmanned vehicle will collide with the obstacle is determined based on a positional relationship between a first coverage area of ​​the obstacle represented by the second detection point and a second coverage area of ​​the unmanned vehicle on the driving trajectory.

[0026] In conjunction with the first aspect, in one possible implementation, determining whether the unmanned vehicle will collide with the obstacle based on the positional relationship between the first coverage area of ​​the obstacle represented by the second detection point and the second coverage area of ​​the unmanned vehicle on the driving trajectory includes:

[0027] Determine a second distance between the second detection point and a corresponding trajectory point of the first detection line segment;

[0028] When the second distances are all less than a second threshold, determining that there is a risk of collision between the unmanned vehicle and the obstacle;

[0029] The second threshold is related to the body width of the unmanned vehicle.

[0030] In conjunction with the first aspect, in one possible implementation, determining whether the unmanned vehicle will collide with the obstacle based on the positional relationship between the first coverage area of ​​the obstacle represented by the second detection point and the second coverage area of ​​the unmanned vehicle on the driving trajectory includes:

[0031] Aggregating the second detection points to determine a second line segment formed by the second detection points;

[0032] Determining a first coverage area representing the obstacle based on the determined positional relationship between the second line segments;

[0033] Determining whether there is spatial overlap between the first coverage area and a second coverage area of ​​the unmanned vehicle on the driving trajectory;

[0034] In the case of spatial overlap, it is determined that there is a risk of collision between the unmanned vehicle and the obstacle.

[0035] In conjunction with the first aspect, in one possible implementation, when there is spatial overlap, determining whether the unmanned vehicle will collide with the obstacle based on the position of the spatial overlap includes:

[0036] In a case where the first detection point spatially overlaps with the obstacle data, determining a third detection point from the first detection points that meets the following conditions: the first detection point spatially overlaps with the obstacle data, and there is a first detection point adjacent to the first detection point that does not spatially overlap with the obstacle data;

[0037] Determining a boundary of a first coverage area of ​​the obstacle according to the third detection point;

[0038] Whether the unmanned vehicle will collide with an obstacle is determined based on a boundary of the first coverage area and a positional relationship between the unmanned vehicle and the second coverage area on the driving trajectory.

[0039] In combination with the first aspect, in a possible implementation manner, the obstacle includes: a static irregular obstacle.

[0040] In conjunction with the first aspect, in one possible implementation, when multiple obstacles are detected, searching the obstacle data in the obstacle data storage space based on the first detection ranges corresponding to the respective trajectory points to determine whether the obstacle data spatially overlaps with any of the first detection ranges includes:

[0041] Searching the obstacle data in the obstacle data storage space to determine the coordinates of the corner points of the geometric figures formed by the areas covered by each obstacle;

[0042] Determine, based on the first detection ranges overlapped by the position coordinates of each corner point, a third distance from the position coordinates of each corner point to the trajectory point corresponding to the overlapped first detection range;

[0043] Determine the shortest distance among the corresponding third distances for each obstacle;

[0044] Sort the shortest distances corresponding to each obstacle in ascending order to obtain the detection order of each obstacle;

[0045] According to the detection order of each obstacle, the corresponding obstacle data is searched in the storage space of the corresponding obstacle data in turn to determine whether the currently detected obstacle data has spatial overlap with any first detection range.

[0046] In a second aspect, an obstacle collision detection device is provided, comprising:

[0047] The obstacle data storage module is used to detect obstacles while the unmanned vehicle is driving and store the detected obstacle data according to a preset data structure;

[0048] an obstacle data search module configured to traverse the trajectory points on the driving trajectory ahead of the unmanned vehicle and, for each trajectory point, determine the normal direction of the driving trajectory at the trajectory point and a first detection range corresponding to the trajectory point along the normal direction; and, based on the first detection range corresponding to each trajectory point, search the obstacle data in the obstacle data storage space to determine whether any obstacle data spatially overlaps with any of the first detection ranges;

[0049] The collision detection module is used to determine whether the unmanned vehicle will collide with an obstacle based on the position of the spatial overlap when there is spatial overlap.

[0050] In a third aspect, an unmanned vehicle is provided, comprising: an obstacle collision detection device as described in the second aspect.

[0051] The beneficial effects of the embodiments of the present disclosure include:

[0052] The obstacle collision detection method, device, and unmanned vehicle provided by the embodiments of the present disclosure include: the unmanned vehicle performs obstacle detection during driving and stores the detected obstacle data according to a preset data structure; traverses the trajectory points on the driving trajectory in front of the unmanned vehicle, and for each trajectory point, determines the normal direction of the driving trajectory at the trajectory point and a first detection range corresponding to the normal direction of the trajectory point; searches the obstacle data in the storage space of the obstacle data based on the first detection range corresponding to each trajectory point, and determines whether there is spatial overlap between the obstacle data and any of the first detection ranges; if there is spatial overlap, determines whether the unmanned vehicle and the obstacle will collide based on the location of the spatial overlap. The obstacle collision detection method provided by the embodiments of the present disclosure constructs a preset data structure to store the obstacle data, constructs a first detection range corresponding to the normal direction of the driving trajectory at each trajectory point on the driving trajectory in front of the unmanned vehicle, and detects whether there is an actual geometric collision between the driving lane of the unmanned vehicle and the irregular obstacle on the driving trajectory. Compared with the existing technology, it can accurately and effectively detect collisions with obstacles of various shapes, while ensuring safety, effectively utilizing the passage space and avoiding vehicle misparking due to erroneous collision detection results. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 One of the flow charts of the obstacle collision detection method provided in the embodiment of the present disclosure;

[0054] Figure 2 A schematic diagram of an embodiment of the present disclosure using a spherical envelope method for obstacle collision detection;

[0055] Figure 3 A schematic diagram of an embodiment of the present disclosure using the AABBbox envelope method for obstacle collision detection;

[0056] Figure 4 One of the schematic diagrams of using the OBBbox envelope method to detect collisions with obstacles provided in an embodiment of the present disclosure;

[0057] Figure 5 The second schematic diagram of using the OBBbox envelope method to detect collisions with obstacles provided in an embodiment of the present disclosure;

[0058] Figure 6 A schematic diagram of an obstacle collision detection method provided by an embodiment of the present disclosure;

[0059] Figure 7 This is a second flow chart of the obstacle collision detection method provided by an embodiment of the present disclosure;

[0060] Figure 8This is a structural diagram of the obstacle collision detection device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0061] The present disclosure provides an obstacle collision detection method, device, and unmanned vehicle. Preferred embodiments of the present disclosure are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are intended only to illustrate and explain the present disclosure and are not intended to limit the present disclosure. Furthermore, the embodiments and features within the embodiments of the present disclosure may be combined with one another unless there is a conflict.

[0062] The present disclosure provides a method for detecting collisions with obstacles. Figure 1 As shown, including:

[0063] S101: The unmanned vehicle detects obstacles while driving and stores the detected obstacle data according to a preset data structure;

[0064] S102, traversing the trajectory points on the driving trajectory in front of the unmanned vehicle, and for each trajectory point, determining the normal direction of the driving trajectory at the trajectory point, and a first detection range corresponding to the trajectory point along the normal direction;

[0065] S103: Search the obstacle data in the obstacle data storage space according to the first detection ranges corresponding to the respective trajectory points to determine whether any obstacle data spatially overlaps with any of the first detection ranges;

[0066] S104: In the case of spatial overlap, determine whether the unmanned vehicle will collide with the obstacle based on the position of the spatial overlap.

[0067] In the disclosed embodiments, the unmanned vehicle can be a mining vehicle equipped for unmanned transportation in mines. In mining environments, unmanned vehicles operate in complex scenarios, often presenting a variety of irregularly shaped obstacles, such as large fallen rocks, earth piles, and retaining walls. Unmanned vehicles are typically equipped with advanced sensors to detect obstacles during operation and generate obstacle data. These sensors may include lidar, cameras, and millimeter-wave radar. For example, a lidar can emit a laser beam and receive the reflected signal, measuring the round-trip time or phase difference to calculate the distance to the obstacle and its location coordinates, generating obstacle data. A camera can capture images and videos of the vehicle's surroundings and, using image processing and deep learning algorithms, identify and classify objects in the images to detect obstacles and generate obstacle data. Millimeter-wave radar utilizes electromagnetic waves in the millimeter-wave frequency band to transmit and receive reflected waves to measure the distance between the unmanned vehicle and obstacles, generating obstacle data. Millimeter-wave radar can assist lidar and cameras in obstacle detection. In the traditional way, after obtaining the obstacle data, the collision detection of obstacles is usually performed by using the sphere envelope, AABBbox envelope, OBBbox envelope and grid map construction method. For example, Figure 2 As shown, the basic idea of ​​the spherical envelope method is to construct the circumscribed circle of the obstacle based on the geometric shape of the obstacle, and to construct the circumscribed circle of the unmanned vehicle based on the geometric shape of the unmanned vehicle. a and c represent unmanned vehicles, and b and d represent obstacles. Then, the distance between the center of the circumscribed circle of the obstacle and the center of the circumscribed circle of the unmanned vehicle, and the sum of the radius of the circumscribed circle of the obstacle and the radius of the circumscribed circle of the unmanned vehicle are calculated respectively. If the distance between the center of the circumscribed circle of the obstacle and the center of the circumscribed circle of the unmanned vehicle is less than the sum of the radius of the circumscribed circle of the obstacle and the radius of the circumscribed circle of the unmanned vehicle, it is determined that a collision occurs between the unmanned vehicle and the obstacle, otherwise no collision occurs. However, if Figure 2 As shown in the figure, the situation where there is no collision between the unmanned vehicle c and the obstacle d will be mistakenly detected as a collision, which will lead to problems such as the vehicle stopping by mistake. Therefore, the use of the spherical envelope method will cause the area where there is no collision to be mistakenly detected as a collision, resulting in the vehicle being unable to pass. Figure 3As shown, the basic idea of ​​the AABBbox envelope method is to construct a plane rectangular coordinate system, construct the circumscribed rectangle of the obstacle (i.e., AABBbox) based on the geometric figure of the obstacle, and construct the circumscribed rectangle of the unmanned vehicle based on the geometric figure of the unmanned vehicle. e and g represent unmanned vehicles, and f and h represent obstacles. Determine whether there is overlap between the projections of the circumscribed rectangle of the obstacle and the circumscribed rectangle of the unmanned vehicle in the x-axis and y-axis directions of the plane rectangular coordinate system. If there is an overlap, it is determined that a collision occurs between the unmanned vehicle and the obstacle, otherwise no collision occurs. For example, the AABBbox envelope method is used to detect the collision between the unmanned vehicle e and the obstacle f, and the detection rate of this method is faster. However, for the overlapping part between the circumscribed rectangle of the unmanned vehicle g and the circumscribed rectangle of the obstacle h, a large area of ​​non-drivable area is generated. For narrow scenes in mines, it will lead to waste of space, which is not conducive to the normal operation of unmanned vehicles. As Figure 4 As shown in the figure, i represents the unmanned vehicle and j represents the obstacle. Compared with the AABBbox envelope method, the bounding rectangle constructed by the OBBbox envelope method can rotate with the direction of the unmanned vehicle or obstacle, and the size of the bounding rectangle will not change. In conjunction with the separating axis theorem, it is more accurate than the AABBbox envelope method to determine whether there is overlap, but the computational complexity is greater. However, as Figure 5 As shown, the OBBbox envelope method cannot produce effective collision detection results for irregular obstacles such as concave polygons. For example, in mining environments, there are numerous irregular concave polygon retaining walls. In the narrow concave polygon areas, the OBBbox envelope method cannot fully utilize the space for collision detection. The grid map method can represent obstacles as scattered grid points, but this method is computationally intensive and lacks tracking properties, resulting in poor compatibility and inconvenience for use with other modules.

[0068] In the disclosed embodiments, the unmanned vehicle detects obstacles while driving. The obstacle data can be point cloud data obtained by scanning the vehicle's direction of travel using a laser radar (LiDAR). This obstacle data includes the location coordinates of the obstacles. The detected obstacle data is stored according to a preset data structure. The preset data structure may include a kd-tree data structure. A kd-tree can be a data structure used to organize multidimensional spatial data, such as point cloud data on a two-dimensional plane or point cloud data in three-dimensional space. A kd-tree is a special binary tree, with each node representing a point in k-dimensional space. During the construction of a kd-tree, axes are selected according to specific rules. For example, in two-dimensional space, the space is first divided by the x-axis, then by the y-axis, and then the space is divided repeatedly. For example, the obstacle data is projected onto the ground to obtain obstacle location coordinate data in two-dimensional space. To construct a kd-tree in two-dimensional space, the x-axis is first selected as the partitioning axis. All points are sorted by their x-coordinates, and the median point is taken as the root node. Then for the left subtree of the root node, select the y-axis as the partition axis, sort the points in the left subtree according to the y-coordinate, take the median point as the root node of the left subtree, and so on. Figure 6 As shown, the unmanned vehicle k is moving forward along the x direction, l represents an obstacle, and the driving track in front of the unmanned vehicle k includes multiple track points. For each track point, the normal direction of the driving track at the track point is determined. For example, at track point B, the normal direction of the driving track at point B is AC. The first detection range along the normal direction corresponding to the track point B can be a line segment AC, or other shapes, such as a quadrilateral area formed by the line segments corresponding to the normal directions of two adjacent track points. For example, Figure 6 The quadrilateral AEFD region in .

[0069] Furthermore, the track points along the unmanned vehicle's forward trajectory are traversed. Based on the first detection ranges corresponding to each track point, the obstacle data storage space is searched for obstacle data to determine whether any obstacle data spatially overlaps with any of the first detection ranges. For example, all obstacle data detected by the unmanned vehicle during driving is stored in a KD tree. Each obstacle data item can be the location coordinates of a point, and the KD tree can efficiently organize this data by recursively partitioning the space. The first detection ranges corresponding to each track point are traversed. For each first detection range corresponding to each track point, a range query is performed in the KD tree to find obstacle data that spatially overlaps with the current first detection range. Starting from the root node of the KD tree, the partitioning dimension of the current node is compared with the first detection range to determine whether the first detection range spatially overlaps with the spatial region partitioned by the current node. If there is spatial overlap, the range query is recursively performed in the left and right subtrees of the current node until a leaf node is reached. During the query, the location coordinates of all points within the first detection range that spatially overlap with the obstacle data are recorded. During the range query process, if at least one obstacle data is found to spatially overlap with the current first detection range, it can be determined that an obstacle exists within the first detection range. If no spatial overlap with the obstacle data is found within all first detection ranges, it can be determined that no obstacle exists within the first detection range of the current trajectory point. In the event of spatial overlap, whether the unmanned vehicle and the obstacle will collide is determined based on the location of the spatial overlap. For example, in the event of spatial overlap, the position coordinates of all recorded points within the first detection range that spatially overlap with the obstacle data can be used to obtain the coverage area of ​​the obstacle. Based on the coverage area of ​​the obstacle and the coverage area of ​​the driving lane on the unmanned vehicle's driving trajectory, it can be determined whether the unmanned vehicle and the obstacle will collide.

[0070] In the embodiment of the present application, various obstacle data are stored by constructing a preset data structure. For the driving trajectory in the driving direction planned by the unmanned vehicle, a first detection range in the normal direction is constructed for each trajectory point on the trajectory. The first detection range can accurately capture the potential risks of the unmanned vehicle in the driving channel, and realize effective collision detection of the obstacle by judging whether there is spatial overlap between the first detection range and the obstacle in a geometric sense. It can accurately adapt to obstacles of different shapes, such as concave polygonal obstacles such as walls, and can perform accurate collision detection. Under the premise of ensuring driving safety, the road traffic space is maximized, and abnormal parking of vehicles due to misjudgment is effectively avoided, which significantly improves the smoothness and reliability of the unmanned vehicle's driving.

[0071] In another embodiment of the present disclosure, in the above step S102, for each trajectory point, determining the normal direction of the driving trajectory at the trajectory point and the first detection range corresponding to the trajectory point along the normal direction includes:

[0072] Step 1: For each trajectory point, determine the normal direction of the driving trajectory at the trajectory point and a first detection line segment along the normal direction of the trajectory point; wherein the length of the first detection line segment is related to a preset detection boundary;

[0073] In step S103, the obstacle data is searched in the obstacle data storage space according to the first detection ranges corresponding to the respective trajectory points to determine whether the obstacle data spatially overlaps with any of the first detection ranges, including:

[0074] Step 2: Discretize the first detection line segments corresponding to the trajectory points to determine a plurality of first detection points located on the first detection line segments;

[0075] Step 3: For each first detection point, based on the hierarchical spatial index structure of the obstacle data, search the obstacle data in the storage space of the obstacle data to determine whether there is spatial overlap between the first detection point and the obstacle data;

[0076] The hierarchical spatial index structure of the obstacle data is related to the preset data structure based on which the obstacles are stored.

[0077] In the embodiment of the present disclosure, by determining the normal direction of the trajectory point and its first detection segment, discretizing the first detection segment to obtain multiple first detection points, and searching obstacle data based on a hierarchical spatial index structure, efficient detection of whether there is spatial overlap between the first detection point and the obstacle data is achieved. Figure 6 As shown, for step 1 above, the direction of the normal perpendicular to the tangent of the driving trajectory at the trajectory point is determined using geometric methods (such as the principle of vector perpendicularity). Starting from the trajectory point, the normal direction is extended in both directions to obtain a first detection segment. For example, at trajectory point B, the normal direction is extended in both directions to obtain first detection segment AC. The length of the first detection segment is determined by a preset detection boundary. The preset detection boundary can be set based on actual application scenarios, such as vehicle speed, safety distance requirements, and vehicle body width, to limit the detection range. For step 2 above, the first detection segment corresponding to each trajectory point is discretized, and multiple first detection points are determined on the first detection segment. The discretization method can use equally spaced sampling to divide the first detection segment into several equal-length subsegments. The endpoints or specific locations of each subsegment serve as the first detection points. Alternatively, multiple first detection points can be set at key locations on the first detection segment based on the required detection accuracy.

[0078] Regarding step 3 above, for each first detection point, the obstacle data is searched in the storage space of the obstacle data based on the hierarchical spatial index structure of the obstacle data to determine whether there is spatial overlap between the first detection point and the obstacle data. The traversal of each first detection point can start from the first detection point at one endpoint of the first detection line segment and end at the first detection point at the other endpoint of the first detection line segment. Alternatively, starting from the first detection point adjacent to the trajectory point on the first detection line segment, the traversal of each first detection point in a direction away from the trajectory point is performed. Based on the hierarchical spatial index structure of the obstacle data, the obstacle data is searched in the storage space of the obstacle data to determine whether there is spatial overlap between each first detection point and the obstacle data. The hierarchical spatial index structure of the obstacle data is related to the preset data structure used for obstacle storage, for example, a kd-tree data structure. The hierarchical spatial index structure of the obstacle data achieves orderly organization and management of the obstacle data by recursively dividing the space into subspaces at different levels. Based on the coordinates of the first detection point, a hierarchical spatial index structure is used to quickly locate the subspace region containing the obstacle. Obstacle data is searched only within this region, rather than traversing all obstacle data, thereby improving search efficiency. Spatial overlap can be determined by determining whether the distance between the coordinates of the first detection point and the coordinates in the obstacle data is less than a preset distance. By setting a first detection line segment along the normal direction for each trajectory point, the area within the preset detection boundary of the driving trajectory can be fully covered, avoiding missing potential obstacles. The hierarchical spatial index structure of obstacle data can quickly locate obstacle data, reduce the search range, and improve detection efficiency. It also adapts to various obstacle shapes and can accurately locate irregular obstacles.

[0079] In another embodiment of the present disclosure, in step 3 above, for each first detection point, based on the hierarchical spatial index structure of the obstacle data, the obstacle data is searched in the storage space of the obstacle data to determine whether there is spatial overlap between the first detection point and the obstacle data, including:

[0080] Step 1: traverse the first detection points of each first detection line segment, and for each first detection point, determine the regional position coordinates of the area covered by the first detection point;

[0081] Step 2: Based on the hierarchical spatial index structure of the obstacle data, a range search is used to search the obstacle data storage space for the location coordinates of the obstacles contained in the obstacle data to determine whether any obstacle data is within the regional location coordinate range;

[0082] Step 3: If there is any spatial overlap between the first detection point and the obstacle data, determine that there is spatial overlap between the first detection point and the obstacle data.

[0083] In the disclosed embodiment, efficient obstacle location is achieved by traversing the first detection points and determining the regional coordinates of their coverage areas. This is then combined with the hierarchical spatial index structure of the obstacle data to perform a range search, ultimately determining whether the first detection points and obstacles spatially overlap. Regarding step 1, the first detection points of each first detection line segment are traversed. For each first detection point, its "coverage area" is defined and the coordinates of the coverage area are determined. Exemplarily, a rectangular area is constructed with each first detection point as the center and the distance between adjacent first detection points as the side length as the coverage area. The coordinates of the four vertices of the rectangular area are used as the regional coordinates of the coverage area. Regarding step 2, utilizing the hierarchical spatial index structure of the obstacle data, using the coordinates of the coverage area of ​​the first detection points as the query condition, a range search is performed to determine whether any obstacle data falls within the regional coordinate range. A range search involves specifying a query condition, typically a rectangular area in multidimensional space, and finding the coordinates of points in the obstacle data that fall within the range. Regarding step 3, if any obstacle data falls within the regional coordinate range, then spatial overlap is determined between the first detection point and the obstacle data. The coordinates of the first detection point with spatial overlap are recorded to locate the obstacle. By determining the coordinates of the area covered by the first detection point, single-point detection is expanded to area detection, which better reflects the spatial occupancy characteristics of actual obstacles and avoids missed detections due to single-point errors.

[0084] In another embodiment of the present disclosure, in step 3 above, for each first detection point, based on the hierarchical spatial index structure of the obstacle data, the obstacle data is searched in the storage space of the obstacle data to determine whether there is spatial overlap between the first detection point and the obstacle data, including:

[0085] Step 1: Traverse the first detection points of each first detection line segment. For each first detection point, search the obstacle data storage space for the position coordinates of the obstacle contained in the obstacle data using a nearest neighbor search method based on the hierarchical spatial index structure of the obstacle data.

[0086] Step 2: When a first distance between the position coordinates of the first detection point and the position coordinates of any obstacle is less than a first threshold, it is determined that there is spatial overlap between the first detection point and the obstacle data.

[0087] In the disclosed embodiment, a nearest neighbor search is performed by traversing the first detection point and combining it with the hierarchical spatial index structure of the obstacle data. Spatial overlap is determined based on the distance between the first detection point and the obstacle's position coordinates, thereby enabling rapid obstacle location. Regarding step 1 above, the first detection points on each first detection line segment are traversed one by one. Using the position coordinates of each first detection point as the query center, the hierarchical spatial index structure of the obstacle data is used to search for the obstacle closest to the position coordinates of the first detection point. The goal of the nearest neighbor search is to find the data point closest to the position coordinates of the first detection point. For example, in a kd-tree data structure, the nearest neighbor search can select the left or right subtree based on the comparison result of the first detection point's position coordinates with the current node in the current dimension, continuing the search until a leaf node is reached. At a leaf node, the distance between the first detection point's position coordinates and the node is calculated, and the node is set as the current nearest neighbor. Starting from the leaf node, the search is performed upwards, checking whether another subtree of each node may contain a closer point. If the distance between the first detection point's position coordinates and the hyperplane containing the current node is less than the distance to the current nearest neighbor, the search is continued in the next subtree. After tracing back to the root node, the current nearest neighbor point is the coordinate of the nearest obstacle position finally searched. For the above step 2, the distance between the position coordinates of the first detection point and the coordinates of the nearest obstacle position searched is calculated as the first distance. If the first distance is less than the preset first threshold, it is determined that there is spatial overlap between the first detection point and the obstacle data. The first threshold can be determined based on the sensor detection error range, such as the ranging error of the lidar. By directly judging whether there is spatial overlap through the first distance and the first threshold, there is no need for complex geometric intersection calculations, which reduces computing overhead and improves real-time performance. The first threshold can be adjusted according to the detection accuracy of the sensor, which is suitable for scenarios with high real-time requirements such as autonomous driving.

[0088] In another embodiment of the present disclosure, in step S104, if there is spatial overlap, determining whether the unmanned vehicle and the obstacle will collide based on the position of the spatial overlap includes:

[0089] Step 1: When the first detection point and the obstacle data are spatially overlapped, determine a second detection point in the first detection point that is spatially overlapped with the obstacle data;

[0090] Step 2: Determine whether the unmanned vehicle will collide with the obstacle based on the positional relationship between the first coverage area of ​​the obstacle represented by the second detection point and the second coverage area of ​​the unmanned vehicle on the driving trajectory.

[0091] In the embodiment of the present disclosure, a second detection point that spatially overlaps with the obstacle data is determined in the first detection point, and the collision risk between the unmanned vehicle and the obstacle is determined based on the positional relationship between the first coverage area of ​​the obstacle represented by the second detection point and the second coverage area of ​​the unmanned vehicle driving channel. For the above step 1, based on determining whether there is spatial overlap between the first detection point and the obstacle data, if there is spatial overlap, all detection points that spatially overlap with the obstacle data are screened out from the first detection point as the second detection point, such as Figure 6 In the example above, points G and H serve as the second detection points. For step 2 above, the area formed by all the second detection points can be used as the first obstacle coverage area. The second coverage area of ​​the unmanned vehicle on its driving trajectory can be calculated based on the unmanned vehicle's physical dimensions (such as length and width) and its driving trajectory. The projection area of ​​the unmanned vehicle's body outline on the trajectory can be used as the second coverage area of ​​the unmanned vehicle on its driving trajectory. Based on the positional relationship between the first coverage area and the second coverage area, it can be determined whether the unmanned vehicle will collide with the obstacle. For example, at a certain moment, if the first coverage area and the second coverage area intersect in space, a collision risk is determined. Combined with the second coverage area of ​​the unmanned vehicle on its driving trajectory, the available space can be maximized, making it suitable for real-time collision warnings with irregular obstacles.

[0092] like Figure 7 As shown, Figure 7 A flowchart of a method for detecting a collision with an obstacle is shown, comprising the following steps:

[0093] S701, obtaining the detected obstacle data and the track points on the front driving track of the unmanned vehicle; proceeding to step S702;

[0094] S702, storing the detected obstacle data according to a preset data structure; proceeding to step S703;

[0095] S703: For each trajectory point, determine the normal direction of the driving trajectory at the trajectory point and the first detection line segment of the trajectory point along the normal direction; then proceed to step S704;

[0096] S704: Discretize the first detection line segments corresponding to the trajectory points to determine a plurality of first detection points located on the first detection line segments; then proceed to step S705;

[0097] S705. For each first detection point, search the obstacle data in the storage space of the obstacle data based on the hierarchical spatial index structure of the obstacle data to determine whether there is spatial overlap between the first detection point and the obstacle data; if so, proceed to step S706; otherwise, proceed to step S701;

[0098] S706: Determine a second detection point in the first detection point that spatially overlaps with the obstacle data; and proceed to step S707.

[0099] S707: Determine whether the unmanned vehicle will collide with the obstacle based on the positional relationship between the first coverage area of ​​the obstacle represented by the second detection point and the second coverage area of ​​the unmanned vehicle on the driving trajectory. This process ends.

[0100] In another embodiment of the present disclosure, in step 2 above, determining whether the unmanned vehicle will collide with the obstacle based on the positional relationship between the first coverage area of ​​the obstacle represented by the second detection point and the second coverage area of ​​the unmanned vehicle on the driving trajectory includes:

[0101] Step 1: Determine a second distance between the second detection point and the corresponding trajectory point of the first detection line segment;

[0102] Step 2: When the second distances are both less than the second threshold, it is determined that there is a risk of collision between the unmanned vehicle and the obstacle;

[0103] The second threshold is related to the body width of the unmanned vehicle.

[0104] In the disclosed embodiment, a second distance between a second detection point and a corresponding trajectory point is calculated, combined with a second threshold value related to the vehicle body width, to determine whether there is a collision risk between the unmanned vehicle and an obstacle. Regarding step one above, the second detection point is located on the first detection line segment. The first detection line segment is an extension of the trajectory point along the normal direction. Therefore, each second detection point is located on a trajectory point corresponding to the first detection line segment. The second distance between the second detection point and the trajectory point corresponding to the first detection line segment is determined. Exemplarily, starting with the first detection point at one end of the first detection line segment and ending with the first detection point at the other end, a search is performed within the obstacle data storage space based on the hierarchical spatial index structure of the obstacle data, and each first detection point is individually determined to determine whether there is spatial overlap with the obstacle data. During the process of individually determining whether there is spatial overlap between each first detection point and the obstacle data, the first detection point that spatially overlaps with the obstacle data is determined as the second detection point. The second distance between the second detection point and the trajectory point corresponding to the first detection line segment is then determined. Regarding step two above, if the second distances are all less than the second threshold value, a collision risk between the unmanned vehicle and the obstacle is determined. The second threshold, related to the vehicle's body width, can be used to measure the second coverage area of ​​the lane along the vehicle's driving trajectory. By directly linking the second distance and the second threshold to the vehicle's physical dimensions, the second coverage area of ​​the lane along the vehicle's driving trajectory is indirectly represented. This method transforms complex spatial relationships into simple numerical comparisons, with low computational overhead and suitable for real-time scenarios. It accurately reflects the proximity of obstacles to the vehicle, thus facilitating collision detection and safe driving in the narrow environments of mines.

[0105] In another embodiment of the present disclosure, in step 2 above, determining whether the unmanned vehicle will collide with the obstacle based on the positional relationship between the first coverage area of ​​the obstacle represented by the second detection point and the second coverage area of ​​the unmanned vehicle on the driving trajectory includes:

[0106] Step 1: Aggregate the second detection points to determine a second line segment formed by the second detection points;

[0107] Step 2: determining a first coverage area representing the obstacle based on the determined positional relationship between the second line segments;

[0108] Step 3: Determine whether there is spatial overlap between the first coverage area and the second coverage area on the driving trajectory of the unmanned vehicle;

[0109] Step 4: In the case of spatial overlap, determine whether there is a collision risk between the unmanned vehicle and the obstacle.

[0110] In the disclosed embodiments, the first coverage area of ​​an obstacle is determined by translating from point to line and from line to surface, and then spatially overlapping with the second coverage area of ​​the unmanned vehicle. This enables collision risk analysis from discrete points to a global area. Regarding step one, for each trajectory point, the second detection points are aggregated. A second line segment formed by these second detection points can be determined by point set fitting or connecting adjacent points. For example, if these second detection points are distributed in an orderly pattern on the first detection line segment, second detection points with similar distances can be connected to form a second detection line segment. Regarding step two, the spatial relationship between the second line segments is analyzed. For example, if multiple second line segments are parallel and adjacent, these second line segments can be used to form the first coverage area representing the obstacle. Regarding step three, the spatial relationship between the first coverage area of ​​the obstacle and the second coverage area of ​​the unmanned vehicle on its trajectory is analyzed to determine whether there is any overlap between them. For example, spatial overlap can be determined for the first and second coverage areas by determining whether edges intersect or whether vertices are within each other's areas. Regarding step four, if there is spatial overlap, a collision risk is determined between the unmanned vehicle and the obstacle. By aggregating the second detection points to form a second line segment and determining the first coverage area, the obstacle's outline and extension direction can be more accurately characterized, ensuring collision judgments are more consistent with the obstacle's actual physical form. Combined with the second coverage area of ​​the autonomous vehicle's trajectory, spatial conflicts between the autonomous vehicle and the obstacle can be determined, making it suitable for real-time obstacle avoidance in complex environments.

[0111] In another embodiment of the present disclosure, in step S104, if there is spatial overlap, determining whether the unmanned vehicle and the obstacle will collide based on the position of the spatial overlap includes:

[0112] Step 1: When a first detection point spatially overlaps with the obstacle data, determine a third detection point from the first detection points that meets the following conditions: the first detection point spatially overlaps with the obstacle data, and a first detection point adjacent to the first detection point does not spatially overlap with the obstacle data.

[0113] Step 2: Determine the boundary of the first coverage area of ​​the obstacle based on the third detection point;

[0114] Step 3: Determine whether the unmanned vehicle will collide with an obstacle based on the boundary of the first coverage area and the positional relationship between the second coverage area on the driving trajectory of the unmanned vehicle.

[0115] In the disclosed embodiment, a third detection point with boundary characteristics is first selected. The boundary of the first coverage area of ​​the obstacle is determined based on the third detection point. Finally, the collision risk is determined based on the positional relationship between the boundary of the first coverage area and the second coverage area of ​​the unmanned vehicle on the driving trajectory, thereby achieving accurate extraction of the obstacle contour and collision analysis. Regarding step 1 above, if there is spatial overlap between the first detection point and the obstacle data, a third detection point is selected from the first detection points. The third detection point has the following characteristics: the third detection point spatially overlaps with the obstacle data, and there is a first detection point adjacent to the third detection point that does not spatially overlap with the obstacle data. In one possible implementation, starting with the first detection point at one end of the first detection line segment and ending with the first detection point at the other end, the obstacle data is searched in the obstacle data storage space based on the hierarchical spatial index structure of the obstacle data, and each first detection point is individually determined to determine whether there is spatial overlap with the obstacle data. During the process of determining whether each first detection point spatially overlaps with the obstacle data, a first detection point that spatially overlaps with the obstacle data is identified, and if the first detection point immediately preceding the first detection point does not spatially overlap with the obstacle data, or the first detection point immediately following the first detection point does not spatially overlap with the obstacle data, then the first detection point is determined as the third detection point. By screening the third detection points for boundary features, the edge of the obstacle on the first detection line segment is precisely located. Regarding step 2 above, the boundary of the obstacle's first coverage area is determined based on the third detection points. The third detection points are arranged in an orderly manner on each first detection line segment. The third detection points are connected in the order in which the first detection line segments corresponding to each trajectory point on the unmanned vehicle's driving trajectory are traversed, thereby determining the boundary of the obstacle's first coverage area. Regarding step 3 above, the positional relationship between the boundary of the obstacle's first coverage area and the second coverage area on the unmanned vehicle's driving trajectory is analyzed to determine whether the unmanned vehicle and the obstacle will collide. For example, determining whether the boundary of the obstacle's first coverage area and the second coverage area geometrically intersect is used to determine whether the unmanned vehicle and the obstacle will collide. By filtering the third detection points, the edge features of obstacles can be accurately captured, avoiding misjudging non-boundary points as contour points, and improving the accuracy of obstacle shape restoration. Constructing the boundary of the first coverage area using only third detection points with boundary features reduces the number of points involved in the calculation, making processing more efficient, and focusing on the key outline information of the obstacle, achieving higher real-time performance. This more accurately reflects the actual spatial conflict between the unmanned vehicle and the obstacle, avoiding unnecessary obstacle avoidance and parking maneuvers due to misjudgment of the obstacle's non-boundary area.

[0116] In yet another embodiment of the present disclosure, the obstacle includes: a static irregular obstacle.

[0117] In the embodiments of the present disclosure, static irregular obstacles may refer to obstacles with fixed positions and shapes and various shapes in the unmanned operation scenario of a mine, such as convex polygonal obstacles and concave polygonal obstacles. For example, large fallen rocks, earth piles, retaining walls, etc. in the unmanned operation scenario of a mine.

[0118] In another embodiment of the present disclosure, when multiple obstacles are detected, in step S103, searching the obstacle data in the obstacle data storage space based on the first detection ranges corresponding to the respective trajectory points to determine whether the obstacle data spatially overlaps with any of the first detection ranges includes:

[0119] Step 1: Search the obstacle data in the obstacle data storage space to determine the corner point coordinates of the geometric figure formed by each obstacle coverage area;

[0120] Step 2: Determine a third distance between each corner point position coordinate and the trajectory point corresponding to the first detection range overlapped by each corner point position coordinate;

[0121] Step 3: Determine the shortest distance among the corresponding third distances for each obstacle;

[0122] Step 4: Sort the shortest distances corresponding to each obstacle in ascending order to obtain the detection order of each obstacle;

[0123] Step 5: According to the detection order of each obstacle, the corresponding obstacle data is searched in the storage space of the corresponding obstacle data in turn to determine whether the currently detected obstacle data has spatial overlap with any first detection range.

[0124] In the disclosed embodiment, the unmanned vehicle performs obstacle detection during driving. If multiple obstacles are detected, the obstacles are ranked, prioritizing the collision risk between the unmanned vehicle and the obstacle with the highest collision risk. Regarding step 1 above, the coordinates of the corner points of the geometric figures formed by the obstacle coverage areas are searched within the obstacle data storage space using a hierarchical spatial index structure (such as a KD tree) and recorded. Obstacle coverage areas are often polygonal, such as convex or concave polygons, and corner points can refer to the vertices of the polygons. If the obstacle coverage area is a curved figure such as a circle, it can be discretized and converted into a polygon, and then the corner point coordinates are searched. Regarding step 2 above, if the coordinates of the corner points of the obstacle overlap with the first detection range, the third distance from each corner point coordinate to the trajectory point corresponding to the overlapping first detection range is determined. Regarding step 3 above, since an obstacle can have multiple corner point coordinates, each corner point coordinate can also have multiple third distances from the trajectory point corresponding to the overlapping first detection range. For multiple obstacles, the shortest of the corresponding third distances is determined for each obstacle. This minimum distance represents the shortest distance from the obstacle to a point on the unmanned vehicle's trajectory. For step 4 above, the shortest distances corresponding to each obstacle are sorted from smallest to largest to determine the obstacle detection order. This allows obstacles to be sorted from highest to lowest risk of collision with the unmanned vehicle. For step 5 above, the corresponding obstacle data is searched sequentially in the corresponding obstacle data storage space according to the obstacle detection order to determine whether the currently detected obstacle data overlaps with any of the first detection ranges. Collision detection between the unmanned vehicle and the obstacle is performed based on the order of highest to lowest risk of collision with the unmanned vehicle. Sorting based on the distance from corner points to trajectory points prioritizes detection of nearby obstacles, conforming to the "near-first" safety principle and improving emergency response efficiency. Sorting by shortest distance allows detection to be terminated early, for example, upon detecting a collision risk with a nearby obstacle. This avoids inefficient calculations for distant obstacles and reduces computing power.

[0125] Based on the same disclosed concept, the embodiments of the present disclosure also provide an obstacle collision detection device and an unmanned vehicle. Since the principles of the problems solved by these devices and the unmanned vehicle are similar to those of the aforementioned obstacle collision detection method, the implementation of the device and the unmanned vehicle can refer to the implementation of the aforementioned method, and the repeated parts will not be repeated.

[0126] The present disclosure provides an obstacle collision detection device, such as Figure 8 including:

[0127] The obstacle data storage module 801 is used for detecting obstacles during the driving process of the unmanned vehicle and storing the detected obstacle data according to a preset data structure;

[0128] The obstacle data search module 802 is configured to traverse the trajectory points on the driving trajectory ahead of the unmanned vehicle and, for each trajectory point, determine the normal direction of the driving trajectory at the trajectory point and the first detection range corresponding to the trajectory point along the normal direction; and, based on the first detection range corresponding to each trajectory point, search the obstacle data in the obstacle data storage space to determine whether any obstacle data overlaps with any of the first detection ranges.

[0129] The collision detection module 803 is used to determine whether the unmanned vehicle will collide with an obstacle based on the position of the spatial overlap when there is spatial overlap.

[0130] In another embodiment of the present disclosure, the obstacle data search module 802 is configured to determine, for each trajectory point, a normal direction of the driving trajectory at the trajectory point and a first detection line segment of the trajectory point along the normal direction; wherein the length of the first detection line segment is related to a preset detection boundary;

[0131] Discretely process the first detection line segments corresponding to the respective trajectory points to determine a plurality of first detection points located on the respective first detection line segments;

[0132] For each first detection point, searching the obstacle data in a storage space of the obstacle data based on the hierarchical spatial index structure of the obstacle data to determine whether there is spatial overlap between the first detection point and the obstacle data;

[0133] The hierarchical spatial index structure of the obstacle data is related to the preset data structure based on which the obstacles are stored.

[0134] In another embodiment of the present disclosure, the obstacle data search module 802 is configured to traverse the first detection points of each first detection line segment and determine, for each first detection point, the regional position coordinates of the area covered by the first detection point;

[0135] Searching the storage space of the obstacle data for the position coordinates of the obstacles contained in the obstacle data using a range search method according to the hierarchical spatial index structure of the obstacle data to determine whether any obstacle data is within the regional position coordinate range;

[0136] If so, it is determined that the first detection point has spatial overlap with the obstacle data.

[0137] In another embodiment of the present disclosure, the obstacle data search module 802 is configured to traverse the first detection points of each first detection line segment and, for each first detection point, search the storage space of the obstacle data for the position coordinates of the obstacle contained in the obstacle data using a nearest neighbor search method based on the hierarchical spatial index structure of the obstacle data.

[0138] When a first distance between the position coordinates of the first detection point and the position coordinates of any obstacle is less than a first threshold, it is determined that there is spatial overlap between the first detection point and the obstacle data.

[0139] In another embodiment of the present disclosure, the collision detection module 803 is configured to, when the first detection point and the obstacle data spatially overlap, determine a second detection point in the first detection point that spatially overlaps with the obstacle data;

[0140] Whether the unmanned vehicle will collide with the obstacle is determined based on a positional relationship between a first coverage area of ​​the obstacle represented by the second detection point and a second coverage area of ​​the unmanned vehicle on the driving trajectory.

[0141] In another embodiment of the present disclosure, the collision detection module 803 is configured to determine a second distance between the second detection point and a corresponding trajectory point of the first detection line segment;

[0142] When the second distances are all less than a second threshold, determining that there is a risk of collision between the unmanned vehicle and the obstacle;

[0143] The second threshold is related to the body width of the unmanned vehicle.

[0144] In another embodiment of the present disclosure, the collision detection module 803 is configured to aggregate the second detection points to determine a second line segment formed by the second detection points;

[0145] Determining a first coverage area representing the obstacle based on the determined positional relationship between the second line segments;

[0146] Determining whether there is spatial overlap between the first coverage area and a second coverage area of ​​the unmanned vehicle on the driving trajectory;

[0147] In the case of spatial overlap, it is determined that there is a risk of collision between the unmanned vehicle and the obstacle.

[0148] In another embodiment of the present disclosure, the collision detection module 803 is configured to, when the first detection point spatially overlaps with the obstacle data, determine, from the first detection points, a third detection point that meets the following conditions: the first detection point spatially overlaps with the obstacle data, and a first detection point adjacent to the first detection point does not spatially overlap with the obstacle data;

[0149] Determining a boundary of a first coverage area of ​​the obstacle according to the third detection point;

[0150] Whether the unmanned vehicle will collide with an obstacle is determined based on a boundary of the first coverage area and a positional relationship between the unmanned vehicle and the second coverage area on the driving trajectory.

[0151] In another embodiment of the present disclosure, the obstacle includes: a static irregular obstacle.

[0152] In another embodiment of the present disclosure, when multiple obstacles are detected, the obstacle data search module 802 is configured to search the obstacle data in the obstacle data storage space to determine the coordinates of the corner points of the geometric figures formed by the areas covered by each obstacle;

[0153] Determine, based on the first detection ranges overlapped by the position coordinates of each corner point, a third distance from the position coordinates of each corner point to the trajectory point corresponding to the overlapped first detection range;

[0154] Determine the shortest distance among the corresponding third distances for each obstacle;

[0155] Sort the shortest distances corresponding to each obstacle in ascending order to obtain the detection order of each obstacle;

[0156] According to the detection order of each obstacle, the corresponding obstacle data is searched in the storage space of the corresponding obstacle data in turn to determine whether the currently detected obstacle data has spatial overlap with any first detection range.

[0157] An embodiment of the present disclosure provides an unmanned vehicle, comprising: an obstacle collision detection device as described in any of the above embodiments.

[0158] Through the above description of the embodiments, those skilled in the art will clearly understand that the embodiments of the present disclosure can be implemented through hardware or through software plus the necessary general-purpose hardware platform. Based on this understanding, the technical solutions of the embodiments of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, or mobile hard drive) and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in the various embodiments of the present disclosure.

[0159] Those skilled in the art will understand that the accompanying drawings are merely schematic diagrams of a preferred embodiment, and the modules or processes in the accompanying drawings are not necessarily required for implementing the present disclosure.

[0160] Those skilled in the art will appreciate that the modules in the devices of the embodiments may be distributed in the devices of the embodiments as described in the embodiments, or may be located in one or more devices different from the embodiments with corresponding changes. The modules of the above embodiments may be combined into one module or further split into multiple submodules.

[0161] The serial numbers of the above-mentioned embodiments of the present disclosure are for description only and do not represent the advantages or disadvantages of the embodiments.

[0162] Obviously, those skilled in the art may make various changes and modifications to the present disclosure without departing from the spirit and scope of the present disclosure. Thus, if these modifications and variations of the present disclosure fall within the scope of the claims of the present disclosure and their equivalents, the present disclosure is intended to include these modifications and variations.

Claims

1. A method for detecting an obstacle collision, characterized in that: include: The unmanned vehicle detects obstacles while driving and stores the detected obstacle data according to a preset data structure; Traversing the trajectory points on the driving trajectory ahead of the unmanned vehicle, and determining, for each trajectory point, a normal direction of the driving trajectory at the trajectory point, and a first detection range corresponding to the trajectory point along the normal direction, including: determining, for each trajectory point, a normal direction of the driving trajectory at the trajectory point, and a first detection line segment of the trajectory point along the normal direction; wherein the length of the first detection line segment is related to a preset detection boundary; Searching the obstacle data in the obstacle data storage space based on the first detection ranges corresponding to the respective trajectory points to determine whether any obstacle data spatially overlaps with any of the first detection ranges, including: discretizing the first detection line segments corresponding to the respective trajectory points to determine a plurality of first detection points located on each first detection line segment; and searching the obstacle data in the obstacle data storage space for each first detection point based on a hierarchical spatial index structure of the obstacle data to determine whether the first detection point spatially overlaps with the obstacle data; wherein the hierarchical spatial index structure of the obstacle data is related to a preset data structure based on which the obstacles are stored; In the case of spatial overlap, whether the unmanned vehicle will collide with the obstacle is determined based on the position of the spatial overlap.

2. The method according to claim 1, wherein The step of searching, for each first detection point, the obstacle data in a storage space of the obstacle data based on the hierarchical spatial index structure of the obstacle data to determine whether the first detection point has spatial overlap with the obstacle data includes: Traversing the first detection points of each first detection line segment, and determining, for each first detection point, the regional position coordinates of the area covered by the first detection point; Searching the storage space of the obstacle data for the position coordinates of the obstacles contained in the obstacle data using a range search method according to the hierarchical spatial index structure of the obstacle data to determine whether any obstacle data is within the regional position coordinate range; If so, it is determined that the first detection point has spatial overlap with the obstacle data.

3. The method according to claim 1, wherein The step of searching, for each first detection point, the obstacle data in a storage space of the obstacle data based on the hierarchical spatial index structure of the obstacle data to determine whether the first detection point has spatial overlap with the obstacle data includes: Traversing the first detection points of each first detection line segment, and for each first detection point, searching the storage space of the obstacle data for the position coordinates of the obstacle contained in the obstacle data using a nearest neighbor search method based on the hierarchical spatial index structure of the obstacle data; When a first distance between the position coordinates of the first detection point and the position coordinates of any obstacle is less than a first threshold, it is determined that there is spatial overlap between the first detection point and the obstacle data.

4. The method according to claim 1, wherein In the case of spatial overlap, determining whether the unmanned vehicle will collide with the obstacle based on the position of the spatial overlap further includes: In a case where the first detection point spatially overlaps with the obstacle data, determining a second detection point among the first detection points that spatially overlaps with the obstacle data; Whether the unmanned vehicle will collide with the obstacle is determined based on a positional relationship between a first coverage area of ​​the obstacle represented by the second detection point and a second coverage area of ​​the unmanned vehicle on the driving trajectory.

5. The method according to claim 4, wherein The determining whether the unmanned vehicle will collide with the obstacle based on a positional relationship between a first coverage area of ​​the obstacle represented by the second detection point and a second coverage area of ​​the unmanned vehicle on the driving trajectory includes: Determine a second distance between the second detection point and a corresponding trajectory point of the first detection line segment; When the second distances are all less than a second threshold, determining that there is a risk of collision between the unmanned vehicle and the obstacle; The second threshold is related to the body width of the unmanned vehicle.

6. The method according to claim 4, wherein The determining whether the unmanned vehicle will collide with the obstacle based on a positional relationship between a first coverage area of ​​the obstacle represented by the second detection point and a second coverage area of ​​the unmanned vehicle on the driving trajectory includes: Aggregating the second detection points to determine a second line segment formed by the second detection points; Determining a first coverage area representing the obstacle based on the determined positional relationship between the second line segments; Determining whether there is spatial overlap between the first coverage area and a second coverage area of ​​the unmanned vehicle on the driving trajectory; In the case of spatial overlap, it is determined that there is a risk of collision between the unmanned vehicle and the obstacle.

7. The method according to claim 1, wherein In the case of spatial overlap, determining whether the unmanned vehicle will collide with the obstacle based on the position of the spatial overlap includes: In a case where the first detection point spatially overlaps with the obstacle data, determining a third detection point from the first detection points that meets the following conditions: the first detection point spatially overlaps with the obstacle data, and there is a first detection point adjacent to the first detection point that does not spatially overlap with the obstacle data; Determining a boundary of a first coverage area of ​​the obstacle according to the third detection point; Whether the unmanned vehicle will collide with an obstacle is determined based on a boundary of the first coverage area and a positional relationship between the unmanned vehicle and the second coverage area on the driving trajectory.

8. The method according to claim 1, wherein The obstacles include: static irregular obstacles.

9. The method according to claim 1, wherein In the case where multiple obstacles are detected, searching the obstacle data in the storage space of the obstacle data according to the first detection ranges corresponding to the respective trajectory points to determine whether the obstacle data spatially overlaps with any of the first detection ranges includes: Searching the obstacle data in the obstacle data storage space to determine the coordinates of the corner points of the geometric figures formed by the areas covered by each obstacle; Determine, based on the first detection ranges overlapped by the position coordinates of each corner point, a third distance from the position coordinates of each corner point to the trajectory point corresponding to the overlapped first detection range; Determine the shortest distance among the corresponding third distances for each obstacle; Sort the shortest distances corresponding to each obstacle in ascending order to obtain the detection order of each obstacle; According to the detection order of each obstacle, the corresponding obstacle data is searched in the storage space of the corresponding obstacle data in turn to determine whether the currently detected obstacle data has spatial overlap with any first detection range.

10. An obstacle collision detection device, characterized in that: include: The obstacle data storage module is used to detect obstacles while the unmanned vehicle is driving and store the detected obstacle data according to a preset data structure; an obstacle data search module configured to traverse the trajectory points on the driving trajectory ahead of the unmanned vehicle and, for each trajectory point, determine the normal direction of the driving trajectory at the trajectory point and a first detection range corresponding to the trajectory point along the normal direction; and, based on the first detection range corresponding to each trajectory point, search the obstacle data in the obstacle data storage space to determine whether any obstacle data spatially overlaps with any of the first detection ranges; A collision detection module is used to determine whether the unmanned vehicle will collide with an obstacle based on the position of the spatial overlap when there is spatial overlap; The obstacle data search module is configured to determine, for each trajectory point, a normal direction of the driving trajectory at the trajectory point and a first detection line segment of the trajectory point along the normal direction; wherein the length of the first detection line segment is related to a preset detection boundary; discretize the first detection line segments corresponding to each trajectory point to determine a plurality of first detection points located on each first detection line segment; and, for each first detection point, search the obstacle data in a storage space of the obstacle data based on a hierarchical spatial index structure of the obstacle data to determine whether there is spatial overlap between the first detection point and the obstacle data; wherein the hierarchical spatial index structure of the obstacle data is related to a preset data structure based on which the obstacles are stored.

11. An unmanned vehicle, characterized in that: The device comprises the obstacle collision detection device as claimed in claim 10.

Images