Methods, devices, equipment, media, and products for positioning underground vehicles

Abstract

This invention discloses a method, apparatus, equipment, medium, and product for locating underground vehicles. The method includes: determining the current candidate pose set corresponding to the vehicle to be located based on the historical candidate pose set corresponding to the vehicle at the previous moment and multiple historical pose changes; performing semantic matching based on roadway semantic features, the semantic map corresponding to the mine roadway where the vehicle is located, and the current candidate pose set to determine the semantically matched pose; performing point cloud matching based on the semantically matched pose, the scanned point cloud data corresponding to the vehicle at the current moment, and the point cloud map corresponding to the mine roadway to determine the point cloud matched pose; and fusing the point cloud matched pose with the pre-determined trajectory prediction pose corresponding to the vehicle at the current moment to obtain the target positioning pose corresponding to the vehicle. This technical solution achieves accurate positioning of underground vehicles based on the integration and collaborative work of multiple vehicle-mounted sensors.

Description

Technical Field

[0001] This invention relates to the field of autonomous driving environmental perception technology, and in particular to a method, device, equipment, medium, and product for locating vehicles in underground mines. Background Technology

[0002] Underground mining areas possess natural advantages for unmanned driving applications, such as fewer personnel and simpler obstacles. Underground mining inherently involves inherent dangers, requiring the deployment of numerous equipment, vehicles, and personnel underground. Various natural or man-made disasters and accidents pose a significant threat to the safety of underground workers. Effective organization and management of underground personnel and vehicles, as well as accident and disaster rescue, require precise location data for these elements. Furthermore, remote and intelligent control of underground equipment relies heavily on real-time, accurate location data. Precise positioning technology has become a crucial technological support for safe production in mines.

[0003] In related technologies, underground vehicle positioning typically involves installing positioning equipment (such as UWB or magnetic nail hardware) in the mine, and then combining the positioning equipment with vehicle sensor data to locate the vehicle. However, this positioning method may suffer from problems such as complex operation, time-consuming and labor-intensive processes, and high installation and maintenance costs. Furthermore, various obstacles exist underground, such as ore, equipment, and pipelines, which may obstruct the scanning path of UWB signals and lidar, resulting in low vehicle positioning accuracy. Summary of the Invention

[0004] This invention provides a method, device, equipment, medium, and product for locating underground vehicles, so as to achieve accurate positioning of underground vehicles based on the integration and collaborative operation of multiple vehicle-mounted sensors.

[0005] According to one aspect of the present invention, an underground vehicle positioning method is provided, the method comprising:

[0006] Based on the set of historical candidate poses of the vehicle to be located at the previous time corresponding to the current time and the predetermined changes in the historical poses of the vehicle to be located at multiple historical times, the set of current candidate poses of the vehicle to be located at the current time is determined.

[0007] The semantic features of the roadway corresponding to the vehicle to be located at the current time are obtained, and semantic matching is performed based on the semantic features of the roadway, the semantic map corresponding to the mine roadway where the vehicle to be located is located, and the current candidate pose set to determine the semantic matching pose of the vehicle to be located at the current time; wherein, the semantic features of the roadway are used to indicate the roadway traffic objects existing on the roadway;

[0008] Point cloud matching is performed based on the semantic matching pose, the scanned point cloud data of the vehicle to be located at the current time, and the point cloud map corresponding to the mine roadway to determine the point cloud matching pose of the vehicle to be located at the current time.

[0009] The point cloud matching pose and the pre-determined trajectory prediction pose of the vehicle to be located at the current time are fused to obtain the target positioning pose of the vehicle to be located at the current time.

[0010] According to another aspect of the present invention, an underground vehicle positioning device is provided, the device comprising:

[0011] The candidate pose set determination module is used to determine the current candidate pose set of the vehicle to be located at the current time based on the historical candidate pose set corresponding to the previous time of the vehicle to be located at the current time and the predetermined historical pose change amount of the vehicle to be located at multiple historical times.

[0012] The semantic matching pose determination module is used to obtain the roadway semantic features corresponding to the vehicle to be located at the current time, and to perform semantic matching based on the roadway semantic features, the semantic map corresponding to the mine roadway where the vehicle to be located is located, and the current candidate pose set to determine the semantic matching pose corresponding to the vehicle to be located at the current time; wherein, the roadway semantic features are used to indicate the roadway traffic objects existing on the roadway;

[0013] The point cloud matching pose determination module is used to perform point cloud matching based on the semantic matching pose, the scanned point cloud data of the vehicle to be located at the current time, and the point cloud map corresponding to the mine roadway, so as to determine the point cloud matching pose of the vehicle to be located at the current time.

[0014] The positioning and pose determination module is used to perform pose fusion between the point cloud matching pose and the pre-determined trajectory prediction pose of the vehicle to be located at the current time, so as to obtain the target positioning pose of the vehicle to be located at the current time.

[0015] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0016] At least one processor; and

[0017] A memory communicatively connected to the at least one processor; wherein,

[0018] The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the downhole vehicle positioning method according to any embodiment of the present invention.

[0019] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the downhole vehicle positioning method according to any embodiment of the present invention.

[0020] According to another aspect of the present invention, a computer program product is provided, the computer program product comprising a computer program that, when executed by a processor, implements the downhole vehicle positioning method according to any embodiment of the present invention.

[0021] The technical solution of this invention determines the current candidate pose set of the vehicle to be located at the current moment by using the historical candidate pose set corresponding to the previous moment of the vehicle to be located at the current moment and the predetermined historical pose change amount of the vehicle to be located at multiple historical moments. Further, it acquires the roadway semantic features corresponding to the vehicle to be located at the current moment, and performs semantic matching based on the roadway semantic features, the semantic map corresponding to the mine roadway where the vehicle to be located is located, and the current candidate pose set to determine the semantic matching pose corresponding to the vehicle to be located at the current moment. Further, it determines the semantic matching pose corresponding to the vehicle to be located at the current moment based on the semantic matching pose, the scanned point cloud data corresponding to the vehicle to be located at the current moment, and the data related to the mine roadway. Point cloud matching is performed on the point cloud map corresponding to the shaft tunnel to determine the point cloud matching pose of the vehicle to be located at the current moment. Furthermore, the point cloud matching pose and the pre-determined trajectory prediction pose of the vehicle to be located at the current moment are fused to obtain the target positioning pose of the vehicle to be located at the current moment. This solves the problems of complex operation, time-consuming and labor-intensive, high cost and low positioning accuracy of underground vehicle positioning methods in related technologies. It realizes the effect of accurate positioning of underground vehicles based on the integration and collaborative work of multiple vehicle-mounted sensors, and achieves the effect of improving vehicle positioning accuracy, positioning stability and reliability while reducing positioning costs.

[0022] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a flowchart of an underground vehicle positioning method according to Embodiment 1 of the present invention;

[0025] Figure 2 This is a flowchart of an underground vehicle positioning method according to Embodiment 2 of the present invention;

[0026] Figure 3 This is a schematic diagram of the structure of an underground vehicle positioning device according to Embodiment 3 of the present invention;

[0027] Figure 4 This is a schematic diagram of the structure of an electronic device that implements the underground vehicle positioning method of this invention. Detailed Implementation

[0028] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0029] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0030] Example 1

[0031] Figure 1This is a flowchart of an underground vehicle positioning method provided in Embodiment 1 of the present invention. This embodiment is applicable to the positioning of underground vehicles. The method can be executed by an underground vehicle positioning device, which can be implemented in hardware and / or software and can be configured in a terminal and / or server. Figure 1 As shown, the method includes:

[0032] S110. Based on the set of historical candidate poses of the vehicle to be located at the previous time and the predetermined changes in historical poses of the vehicle to be located at multiple historical times, determine the set of current candidate poses of the vehicle to be located at the current time.

[0033] The vehicle to be located can be a vehicle located underground in a mine. In this embodiment, the vehicle to be located is a vehicle located in a mine roadway, i.e., a vehicle traveling in a mine roadway. The vehicle to be located can be any type of vehicle, optionally, an unmanned vehicle or an autonomous vehicle, etc. The current time can be the time when the vehicle is to be located. The historical candidate pose set can include multiple historical candidate poses corresponding to the vehicle to be located at the previous time. The historical candidate poses can be used to characterize the poses that the vehicle to be located might have existed at the previous time. The historical candidate poses can be determined based on the historical candidate poses corresponding to the previous time and the historical pose changes at multiple historical times. It should be noted that the multiple initial candidate poses corresponding to the vehicle to be located at the initial time can be determined based on the starting position and initial speed of the vehicle to be located at the initial time and the environmental information of the mine where the vehicle to be located is located. The historical time can be any historical time before the current time. The historical pose changes can be used to characterize the pose change information of the vehicle to be located at the historical time. The historical pose changes can include historical position changes and historical attitude changes. The current candidate pose set can include multiple current candidate poses corresponding to the vehicle to be located at the current moment. A current candidate pose can be understood as the pose that the vehicle to be located may have at the current moment. These multiple current candidate poses can include the poses required for the localization process of the vehicle at the current moment.

[0034] In this embodiment, the historical position change can be determined based on the vehicle speed information of the vehicle to be located. Specifically, the historical position change at a given historical moment can be determined based on the historical speed at that moment, the historical speed of the previous moment, and the time interval between the two moments. More specifically, the historical position change at that moment can be obtained by integrating the historical speed at the previous moment and the historical speed of the previous moment based on the time interval. The historical attitude change can be determined based on the angular velocity information of the vehicle to be located. Specifically, the historical attitude change at a given historical moment can be determined based on the historical angular velocity at that moment, the historical angular velocity of the previous moment, and the time interval between the two moments. More specifically, the historical attitude change at that moment can be obtained by integrating the historical angular velocity at the previous moment and the historical angular velocity of the previous moment based on the time interval. It should be noted that the historical angular velocity can be obtained from inertial measurement unit (IMU) data.

[0035] In this embodiment, in order to determine the positioning pose of the vehicle to be located at the current moment, the current candidate pose set of the vehicle to be located at the current moment can be determined first. Then, the positioning pose of the vehicle to be located at the current moment can be determined based on the multiple current candidate poses included in the current candidate pose set. The current candidate pose set of the vehicle to be located at the current moment can be determined based on the historical candidate pose set of the vehicle to be located at the previous moment and the historical pose change amounts at multiple historical moments.

[0036] Optionally, based on the set of historical candidate poses corresponding to the vehicle to be located at the previous time and the predetermined changes in historical poses corresponding to the vehicle at multiple historical times, the current candidate pose set corresponding to the vehicle to be located at the current time is determined. This includes: obtaining the set of historical candidate poses corresponding to the vehicle to be located at the previous time and the changes in historical poses corresponding to multiple historical times; for each historical candidate pose in the set of historical candidate poses, determining the product between the historical candidate pose and multiple historical pose changes to obtain the current candidate pose corresponding to the current time; and determining the current candidate pose set corresponding to the vehicle to be located at the current time based on the multiple current candidate poses corresponding to the current time.

[0037] As an optional implementation of this embodiment, the set of historical candidate poses corresponding to the previous time step of the vehicle to be located and the changes in historical poses corresponding to multiple historical time steps can be obtained. Further, for each historical candidate pose in the set of historical candidate poses, the product between the historical candidate pose and multiple changes in historical poses can be determined, and the resulting product can be used as the current candidate pose corresponding to the current time step. Further, a set of current poses corresponding to the vehicle to be located at the current time step can be constructed based on the multiple current candidate poses corresponding to the current time steps.

[0038] S120. Obtain the semantic features of the roadway corresponding to the vehicle to be located at the current time, and perform semantic matching based on the semantic features of the roadway, the semantic map corresponding to the mine roadway where the vehicle to be located is located, and multiple current candidate pose sets to determine the semantic matching pose corresponding to the vehicle to be located at the current time.

[0039] Semantic features of mine roadways are used to indicate the traffic objects present in the roadways. These features can be understood as characteristics that reflect the basic attributes of the roadways, aiding in subjective understanding and recognition. Optionally, semantic features can include roadway markings, traffic signs, roadway boundaries, signal lights, etc. Mine roadways are a general term for various underground spaces excavated in different rocks along different directions, at different angles, with different cross-sections and lengths, serving different ranges and for different purposes. Semantic maps are typically a high-level map representation. They contain not only geometric information about geographic location but also semantic information about various objects, roadways, traffic rules, etc., in the environment. Generally, environmental data of mine roadways can be collected using sensors installed on vehicles to obtain 3D data, image data, and location data of traffic objects within the roadways. Furthermore, semantic maps can be constructed from the acquired environmental data of the mine roadways, and semantic features can be annotated on the constructed semantic map to mark the traffic objects within the mine roadways on the map. Furthermore, the annotated semantic map can be used as the semantic map corresponding to the mine roadway. The semantic matching pose can be understood as the matching pose determined through a semantic matching operation. It should be noted that the semantic matching pose can be one of multiple current candidate poses included in the current candidate pose set.

[0040] In this embodiment, the semantic features of the roadway corresponding to the vehicle to be located at the current moment can be determined based on the roadway image of the mine roadway where the vehicle to be located is located at the current moment.

[0041] Optionally, obtaining the semantic features of the roadway corresponding to the vehicle to be located at the current time includes: acquiring images of the mine roadway where the vehicle to be located is located by a camera device pre-set on the vehicle to be located, so as to obtain the roadway image corresponding to the vehicle to be located at the current time; extracting semantic features from the roadway image to obtain the semantic features of the roadway corresponding to the vehicle to be located at the current time.

[0042] The camera device can be a vehicle-mounted camera. The alleyway image can be an image reflecting the surrounding environment of the vehicle to be located. In this embodiment, semantic feature extraction from the alleyway image can be implemented in various ways. Optionally, semantic feature extraction from the alleyway image can be performed based on a visual semantic segmentation algorithm; semantic feature extraction can be performed based on a semantic segmentation model, etc.

[0043] As an optional implementation of this embodiment, during the movement of the vehicle to be located, images of the mine roadway where the vehicle is located can be acquired using a camera device pre-installed on the vehicle. This yields an image of the roadway corresponding to the vehicle at the current moment. Furthermore, the roadway image can be input into a pre-trained semantic segmentation model to extract features from the roadway image based on the semantic segmentation model, thereby obtaining the semantic features of the roadway corresponding to the vehicle at the current moment.

[0044] In this embodiment, after obtaining the semantic features of the mine roadway corresponding to the vehicle to be located at the current time, a semantic map corresponding to the mine roadway where the vehicle to be located is located can be obtained. Then, semantic matching can be performed based on the roadway semantic features, semantic feature elements in the semantic map, and the current candidate pose set. Furthermore, based on the semantic matching result, the semantic matching pose corresponding to the vehicle to be located at the current time can be determined from multiple current candidate poses in the current candidate pose set.

[0045] S130. Based on the semantic matching pose, the scanned point cloud data corresponding to the vehicle to be located at the current time, and the point cloud map corresponding to the mine roadway, point cloud matching is performed to determine the point cloud matching pose corresponding to the vehicle to be located at the current time.

[0046] In this context, the scanned point cloud data can be understood as point cloud data obtained by scanning the surrounding environment of a vehicle using an onboard radar device. In this embodiment, the scanned point cloud data can be point cloud data obtained by scanning the surrounding environment of the vehicle to be located using an onboard radar device pre-installed on the vehicle to be located. Generally, when a vehicle to be located, equipped with an onboard radar device, is traveling in a mine tunnel, the onboard radar device can scan the surrounding environment of the vehicle to be located. This allows the acquisition of point cloud data of the surrounding environment of the vehicle to be located, which can then be used as the scanned point cloud data. The point cloud map can be a pre-constructed 3D point cloud map of the mine tunnel. Generally, environmental point cloud data of the mine tunnel can be acquired, and feature information from the environmental point cloud data can be extracted in a timely manner using SLAM. Then, a 3D point cloud map of the mine tunnel is constructed based on the feature information. The environmental point cloud data is the perception information of the surrounding environment obtained by sensors (such as LiDAR, cameras, inertial measurement units, etc.) mounted on mobile devices or robots. The point cloud matching pose can be understood as the matching pose determined through point cloud matching operations.

[0047] In this embodiment, given the semantic matching pose, it can be used as an input parameter in the point cloud matching operation. Furthermore, the scanned point cloud data corresponding to the vehicle to be located at the current moment and the point cloud map corresponding to the mine roadway where the vehicle is located can be obtained. Then, point cloud matching can be performed based on the semantic matching pose, the scanned point cloud data, and the point cloud map to obtain the point cloud matching pose corresponding to the vehicle to be located at the current moment.

[0048] Optionally, point cloud matching is performed based on the semantic matching pose, the scanned point cloud data corresponding to the vehicle to be located at the current time, and the point cloud map corresponding to the mine roadway to determine the point cloud matching pose corresponding to the vehicle to be located. This includes: determining the product between the semantic matching pose and a pre-determined transformation parameter matrix to obtain the point cloud pose to be processed; determining the matching error based on the point cloud pose to be processed, the scanned point cloud data, and the actual point cloud data on the point cloud map corresponding to the scanned point cloud data; adjusting the point cloud pose to be processed with the minimum matching error as the optimization objective, and taking the point cloud pose to be processed corresponding to the minimum matching error as the point cloud matching pose corresponding to the vehicle to be located.

[0049] The transformation parameter matrix is ​​used to characterize the transformation relationship between the LiDAR coordinate system and the camera coordinate system. It typically describes how points in one coordinate system are transformed to another, thus achieving data fusion between the LiDAR and the camera. The transformation parameter matrix can include rotation and translation transformation parameters. Generally, the transformation parameter matrix can be determined by calibrating the extrinsic parameters of the LiDAR and camera. Furthermore, the transformation parameter matrix can be determined based on the extrinsic parameter calibration results. In this embodiment, the semantic matching pose is obtained by matching the alleyway image acquired by the vehicle-mounted camera with the semantic map, corresponding to the pose information in the camera coordinate system. Therefore, multiplying the semantic matching pose with the transformation parameter matrix to obtain the point cloud pose to be processed can be understood as transforming the pose in the camera coordinate system to the pose in the LiDAR coordinate system. The matching error can be used to characterize the degree of matching difference between the scanned point cloud data and the point cloud map. The point cloud matching algorithm is mainly used to align two or more frames of point cloud data to obtain their relative pose or transformation matrix. Optionally, point cloud matching algorithms include point-to-surface point cloud matching algorithm (Iterative Closest Point, ICP) and normal vector iterative closest point algorithm (Normal Iterative Closest Point, NICP), etc.

[0050] As an optional implementation of this embodiment, the onboard radar and onboard camera devices of the vehicle to be located can be calibrated using extrinsic parameters to obtain a transformation parameter matrix characterizing the transformation relationship between the lidar coordinate system and the camera coordinate system. Further, given the semantic matching pose, the product between the semantic matching pose and the transformation parameter matrix can be determined, and the resulting product can be used as the point cloud pose to be processed corresponding to the vehicle to be located at the current moment. Further, the point cloud pose to be processed, the scanned point cloud data, and the actual point cloud data corresponding to the scanned point cloud data on the point cloud map can be processed according to the error function corresponding to the preset point cloud matching algorithm to obtain the matching error. Further, the point cloud data to be processed can be adjusted with minimizing the matching error as the optimization objective. Therefore, the point cloud pose to be processed corresponding to the minimum matching error can be used as the point cloud matching pose of the vehicle to be located at the current moment.

[0051] For example, the matching error can be represented by the following formula:

[0052]

[0053] Where E represents the matching error; p i T represents the i-th point in the scanned point cloud data; lm Indicates the pose of the point cloud to be processed; q i This indicates that the actual scan data on the point cloud map is related to p.i The corresponding point; λ represents the regularization coefficient, which is optional, λ∈[0.35,0.55].

[0054] S140. Perform pose fusion between the point cloud matching pose and the pre-determined trajectory prediction pose of the vehicle to be located at the current time to obtain the target positioning pose of the vehicle to be located at the current time.

[0055] The trajectory prediction pose can be understood as the vehicle pose calculated using trajectory extrapolation. The target localization pose can be used to characterize the positioning information of the vehicle to be located at the current moment. Generally, the target localization pose can be used to locate the vehicle, so as to obtain the vehicle's position and driving status in a timely manner. The target localization pose can at least include the vehicle's position and driving status.

[0056] In this embodiment, the predicted trajectory pose of the vehicle to be located at the current moment can be determined based on the historical actual pose of the vehicle to be located at the previous moment. Optionally, the method for determining the predicted trajectory pose includes: obtaining the historical actual pose of the vehicle to be located at the previous moment and the change in the current pose at the current moment; determining the product between the historical actual pose and the change in the current pose to obtain the predicted trajectory pose of the vehicle to be located at the current moment.

[0057] The historical actual pose can be the actual vehicle pose of the vehicle to be located at the previous moment. The current pose change can be used to characterize the pose change from the previous moment to the current moment. The current pose change can include the current position change and the current attitude change.

[0058] In this embodiment, the current vehicle speed of the vehicle to be located at the current moment and the historical vehicle speed at the previous moment can be obtained. Furthermore, the time interval between the current moment and the previous moment can be determined, and the current vehicle speed and the historical vehicle speed can be integrated based on the time interval to obtain the current position change corresponding to the current moment. Also, the current vehicle angular velocity of the vehicle to be located at the current moment and the historical vehicle angular velocity at the previous moment can be obtained using an inertial measurement unit. Furthermore, the time interval between the current moment and the previous moment can be determined, and the current vehicle angular velocity and the historical vehicle angular velocity can be integrated based on the time interval to obtain the current attitude change corresponding to the current moment.

[0059] As an optional implementation of this embodiment, the current vehicle speed of the vehicle to be located at the current moment and the historical vehicle speed at the previous moment can be obtained, and the change in the current position of the vehicle to be located at the current moment can be determined based on the current vehicle speed and the historical vehicle speed; and the current angular velocity of the vehicle to be located at the current moment and the historical vehicle angular velocity at the previous moment can be obtained, and the change in the current attitude of the vehicle to be located at the current moment can be determined based on the current vehicle angular velocity and the historical vehicle angular velocity. Furthermore, the change in current position and the change in current attitude can be used as the change in current pose of the vehicle to be located at the current moment. Further, the historical actual pose of the vehicle to be located at the previous moment and the change in current pose at the current moment can be obtained. Further, the product between the historical actual pose and the change in current pose can be determined, and this product can be used as the predicted trajectory pose of the vehicle to be located at the current moment.

[0060] In practical applications, the predicted trajectory pose can often be directly used as the vehicle's localization pose. However, the predicted trajectory pose may have low accuracy, which in turn leads to low localization accuracy of the vehicle.

[0061] To address this situation, this embodiment utilizes a fusion positioning method to combine matching poses from multiple scenarios. The resulting fused pose can then be used as the target positioning pose for the vehicle at the current moment.

[0062] Optionally, the point cloud matching pose and the pre-determined trajectory prediction pose of the vehicle to be located at the current time are fused to obtain the target positioning pose of the vehicle to be located at the current time, including: interpolating and fusing the point cloud matching pose and the trajectory prediction pose according to the fusion positioning algorithm, and using the fused pose as the target positioning pose of the vehicle to be located at the current time.

[0063] Among them, the fusion localization algorithm can be an algorithm used to fuse at least two poses to obtain the final localization pose.

[0064] As an optional implementation of this embodiment, the point cloud matching pose and trajectory prediction pose of the vehicle to be located at the current moment can be interpolated and fused according to the fusion localization algorithm, and the fused pose can be output. Then, the fused pose can be used as the target localization pose of the vehicle to be located at the current moment.

[0065] The technical solution of this invention determines the current candidate pose set of the vehicle to be located at the current moment by using the historical candidate pose set corresponding to the previous moment of the vehicle to be located at the current moment and the predetermined historical pose change amount of the vehicle to be located at multiple historical moments. Further, it acquires the roadway semantic features corresponding to the vehicle to be located at the current moment, and performs semantic matching based on the roadway semantic features, the semantic map corresponding to the mine roadway where the vehicle to be located is located, and the current candidate pose set to determine the semantic matching pose corresponding to the vehicle to be located at the current moment. Further, it determines the semantic matching pose corresponding to the vehicle to be located at the current moment based on the semantic matching pose, the scanned point cloud data corresponding to the vehicle to be located at the current moment, and the data related to the mine roadway. Point cloud matching is performed on the point cloud map corresponding to the shaft tunnel to determine the point cloud matching pose of the vehicle to be located at the current moment. Furthermore, the point cloud matching pose and the pre-determined trajectory prediction pose of the vehicle to be located at the current moment are fused to obtain the target positioning pose of the vehicle to be located at the current moment. This solves the problems of complex operation, time-consuming and labor-intensive, high cost and low positioning accuracy of underground vehicle positioning methods in related technologies. It realizes the effect of accurate positioning of underground vehicles based on the integration and collaborative work of multiple vehicle-mounted sensors, and achieves the effect of improving vehicle positioning accuracy, positioning stability and reliability while reducing positioning costs.

[0066] Example 2

[0067] Figure 2 This is a flowchart of an underground vehicle positioning method provided in Embodiment 2 of the present invention. Based on the aforementioned embodiments, the method for determining the semantic matching pose is further refined. Optionally, semantic matching is performed based on the semantic features of the mine roadway, the semantic map corresponding to the mine roadway where the vehicle to be located is located, and the current candidate pose set to determine the semantic matching pose of the vehicle to be located at the current moment. This includes: for each current candidate pose in the current candidate pose set, determining the target weight corresponding to the target candidate pose based on the target candidate pose, the mine roadway semantic features, and the semantic map; and determining the semantic matching pose of the vehicle to be located at the current moment from the multiple current candidate poses based on the target weights corresponding to the multiple current candidate poses. For specific implementation details, please refer to the technical solution of this embodiment. Technical terms that are the same as or similar to those in the above embodiments will not be repeated here.

[0068] like Figure 2 As shown, the method includes:

[0069] S210. Based on the set of historical candidate poses of the vehicle to be located at the previous time and the predetermined changes in historical poses of the vehicle to be located at multiple historical times, determine the set of current candidate poses of the vehicle to be located at the current time.

[0070] S220. Obtain the lane semantic features corresponding to the vehicle to be located at the current time. For each current candidate pose in the current candidate pose set, determine the target weight corresponding to the current candidate pose based on the current candidate pose, lane semantic features and semantic map.

[0071] The target weight can be a numerical value that represents the probability of the current candidate pose. Generally, the higher the target weight, the higher the probability that the candidate pose corresponding to that target weight is the final desired matching pose.

[0072] In this embodiment, after obtaining the current candidate pose set corresponding to the vehicle to be located at the current time and the lane semantic features corresponding to the vehicle to be located at the current time, for each current candidate pose in the current candidate pose set, the matching distance between the lane semantic features and the semantic feature elements in the semantic map corresponding to the lane semantic features is determined based on the current candidate pose, the lane semantic features, and the semantic map. Furthermore, the target weight corresponding to the current candidate pose can be determined based on this matching distance.

[0073] Optionally, based on the current candidate pose, lane semantic features, and semantic map, the target weight corresponding to the current candidate pose is determined, including: determining the product between the current candidate pose and the lane semantic features to obtain the visual detection features corresponding to the current candidate pose; determining the distance between the visual detection features and the semantic feature elements in the semantic map corresponding to the lane semantic features to obtain the semantic matching distance corresponding to the current candidate pose; and determining the target weight corresponding to the current candidate pose based on the semantic matching distance.

[0074] As an optional implementation of this embodiment, for each current candidate pose in the current candidate pose set, the product between the current candidate pose and the lane semantic features can be determined, and the resulting product can be used as the visual detection feature corresponding to the current candidate pose. Further, semantic feature elements corresponding to the lane semantic features in the semantic map can be determined, and the distance between the visual detection feature and the semantic feature elements can be determined, and the resulting distance can be used as the semantic matching distance corresponding to the current candidate pose. Further, the semantic matching distance can be processed according to a Gaussian function to obtain the target weight corresponding to the current candidate pose.

[0075] For example, the semantic matching distance can be determined based on the following formula:

[0076]

[0077] Among them, dist i T represents the semantic matching distance corresponding to the i-th current candidate pose; cm Indicates the current candidate pose; s jRepresents the semantic features of the j-th alleyway; m j Indicates the semantic map with s j Corresponding semantic feature elements.

[0078] For example, the target weight can be represented based on the following formula:

[0079]

[0080] in, The target weight corresponding to the i-th current candidate pose is represented by μ; the mean of the semantic matching distance is represented by σ; and the standard deviation of the semantic matching distance is represented by dist. i It represents the semantic matching distance corresponding to the i-th current candidate pose.

[0081] S230. Based on the target weights corresponding to multiple current candidate poses, determine the semantic matching pose of the vehicle to be located at the current moment from multiple current candidate poses.

[0082] In this embodiment, given the target weights corresponding to each current candidate pose, the semantic matching pose of the vehicle to be located at the current moment can be determined from multiple current candidate poses based on the multiple target weights.

[0083] Optionally, based on the target weights corresponding to multiple current candidate poses, the semantic matching pose of the vehicle to be located at the current time is determined from the multiple current candidate poses, including: determining the target weight with the largest weight value based on multiple target weights, and using the current candidate pose corresponding to the determined target weight as the semantic matching pose of the vehicle to be located at the current time.

[0084] As an optional implementation of this embodiment, the target weight with the largest weight value can be determined from multiple target weights. Furthermore, the current candidate pose corresponding to the determined target weight can be used as the semantic matching pose of the vehicle to be located at the current moment.

[0085] S240. Based on the semantic matching pose, the scanned point cloud data corresponding to the vehicle to be located at the current time, and the point cloud map corresponding to the mine roadway, point cloud matching is performed to determine the point cloud matching pose corresponding to the vehicle to be located at the current time.

[0086] S250. The point cloud matching pose and the pre-determined trajectory prediction pose of the vehicle to be located at the current time are fused to obtain the target positioning pose of the vehicle to be located at the current time.

[0087] The technical solution of this invention obtains the roadway semantic features corresponding to the vehicle to be located at the current time. For each current candidate pose in the current candidate pose set, a target weight corresponding to the current candidate pose is determined based on the target candidate pose, the roadway semantic features, and the semantic map. Furthermore, based on the target weights corresponding to multiple current candidate poses, the semantic matching pose corresponding to the vehicle to be located at the current time is determined from multiple current candidate poses. This achieves the effect of accurately locating underground vehicles by extracting and matching semantic features in images. These semantic features are relatively stable in the underground mining environment and are not easily affected by environmental changes, thus improving the stability and accuracy of the positioning.

[0088] Example 3

[0089] Figure 3 This is a structural schematic diagram of an underground vehicle positioning device provided in Embodiment 3 of the present invention. Figure 3 As shown, the device includes: a candidate pose set determination module 310, a semantic matching pose determination module 320, a point cloud matching pose determination module 330, and a localization pose determination module 340.

[0090] The candidate pose set determination module 310 is used to determine the current candidate pose set of the vehicle to be located at the current time based on the historical candidate pose set corresponding to the previous time of the vehicle to be located at the current time and the pre-determined historical pose change amount corresponding to the vehicle to be located at multiple historical time points. The semantic matching pose determination module 320 is used to obtain the roadway semantic features corresponding to the vehicle to be located at the current time, and perform semantic matching based on the roadway semantic features, the semantic map corresponding to the mine roadway where the vehicle to be located is located, and the current candidate pose set to determine the current candidate pose set of the vehicle to be located at the current time. Semantic matching pose; wherein, the roadway semantic features are used to indicate the roadway traffic objects existing on the roadway; point cloud matching pose determination module 330 is used to perform point cloud matching based on the semantic matching pose, the scanned point cloud data of the vehicle to be located at the current time, and the point cloud map corresponding to the mine roadway, so as to determine the point cloud matching pose of the vehicle to be located at the current time; positioning pose determination module 340 is used to perform pose fusion of the point cloud matching pose and the pre-determined trajectory prediction pose of the vehicle to be located at the current time, so as to obtain the target positioning pose of the vehicle to be located at the current time.

[0091] The technical solution of this invention determines the current candidate pose set of the vehicle to be located at the current moment by using the historical candidate pose set corresponding to the previous moment of the vehicle to be located at the current moment and the predetermined historical pose change amount of the vehicle to be located at multiple historical moments. Further, it acquires the roadway semantic features corresponding to the vehicle to be located at the current moment, and performs semantic matching based on the roadway semantic features, the semantic map corresponding to the mine roadway where the vehicle to be located is located, and the current candidate pose set to determine the semantic matching pose corresponding to the vehicle to be located at the current moment. Further, it determines the semantic matching pose corresponding to the vehicle to be located at the current moment based on the semantic matching pose, the scanned point cloud data corresponding to the vehicle to be located at the current moment, and the data related to the mine roadway. Point cloud matching is performed on the point cloud map corresponding to the shaft tunnel to determine the point cloud matching pose of the vehicle to be located at the current moment. Furthermore, the point cloud matching pose and the pre-determined trajectory prediction pose of the vehicle to be located at the current moment are fused to obtain the target positioning pose of the vehicle to be located at the current moment. This solves the problems of complex operation, time-consuming and labor-intensive, high cost and low positioning accuracy of underground vehicle positioning methods in related technologies. It realizes the effect of accurate positioning of underground vehicles based on the integration and collaborative work of multiple vehicle-mounted sensors, and achieves the effect of improving vehicle positioning accuracy, positioning stability and reliability while reducing positioning costs.

[0092] Optionally, the historical candidate pose set includes multiple historical candidate poses corresponding to the vehicle to be located at the previous moment; the candidate pose set determination module 310 includes: a historical pose acquisition unit, a current candidate pose determination unit, and a candidate pose set determination unit.

[0093] The historical pose acquisition unit is used to acquire the set of historical candidate poses of the vehicle to be located at the current moment and the changes in historical poses at multiple historical moments.

[0094] The current candidate pose determination unit is used to determine the product between the historical candidate pose and multiple historical pose transformation quantities for each historical candidate pose in the historical candidate pose set, so as to obtain the current candidate pose corresponding to the current moment.

[0095] The candidate pose set determination unit is used to determine the current candidate pose set of the vehicle to be located at the current time based on multiple current candidate poses corresponding to the current time.

[0096] Optionally, the semantic matching pose determination module 320 includes: a tunnel image acquisition unit and a semantic feature extraction unit.

[0097] The tunnel image acquisition unit is used to acquire images of the mine tunnel where the vehicle to be located is located by a camera device pre-set on the vehicle to be located, so as to obtain the tunnel image corresponding to the vehicle to be located at the current time.

[0098] The semantic feature extraction unit is used to extract semantic features from the lane image to obtain the lane semantic features corresponding to the vehicle to be located at the current time.

[0099] Optionally, the semantic matching pose determination module 320 includes: a pose weight determination unit and a semantic matching pose determination unit.

[0100] The pose weight determination unit is used to determine the target weight corresponding to each current candidate pose in the current candidate pose set, based on the target candidate pose, the alleyway semantic features, and the semantic map.

[0101] The semantic matching pose determination unit is used to determine the semantic matching pose of the vehicle to be located at the current time from the multiple current candidate poses according to the target weights corresponding to the multiple current candidate poses.

[0102] Optionally, the pose weight determination unit includes: a visual feature determination subunit, a semantic matching distance determination subunit, and a pose weight determination subunit.

[0103] A visual feature determination subunit is used to determine the product between the current candidate pose and the alleyway semantic features to obtain the visual detection features corresponding to the current candidate pose.

[0104] The semantic matching distance determination subunit is used to determine the distance between the visual detection feature and the semantic feature element in the semantic map that corresponds to the alleyway semantic feature, so as to obtain the semantic matching distance corresponding to the current candidate pose;

[0105] The pose weight determination subunit is used to determine the target weight corresponding to the current candidate pose based on the semantic matching distance.

[0106] Optionally, the semantic matching pose determination unit is specifically used to determine the target weight with the largest weight value based on multiple target weights, and to use the current candidate pose corresponding to the determined target weight as the semantic matching pose corresponding to the vehicle to be located.

[0107] Optionally, the point cloud matching pose determination module 330 includes: a point cloud pose determination unit, a matching error determination unit, and a point cloud matching pose determination unit.

[0108] A point cloud pose determination unit is used to determine the product between the semantic matching pose and a pre-determined transformation parameter matrix to obtain the point cloud pose to be processed.

[0109] The matching error determination unit is used to process the pose of the point cloud to be processed, the scanned point cloud data, and the actual point cloud data on the point cloud map corresponding to the scanned point cloud data according to a preset point cloud matching algorithm, so as to obtain the matching error.

[0110] The point cloud matching pose determination unit is used to adjust the pose of the point cloud to be processed with the minimum matching error as the optimization target, and to take the point cloud pose to be processed corresponding to the minimum matching error as the point cloud matching pose of the vehicle to be located.

[0111] Optionally, the positioning pose determination module 340 is specifically used to interpolate and fuse the point cloud matching pose and the trajectory prediction pose according to the fusion positioning algorithm, and use the fused pose as the target positioning pose of the vehicle to be located at the current time.

[0112] Optionally, the device further includes: a historical pose acquisition module and a trajectory prediction pose determination module.

[0113] The historical pose acquisition module is used to acquire the historical actual pose of the vehicle to be located at the previous moment and the change in the current pose at the current moment.

[0114] The trajectory prediction pose determination module is used to determine the product between the historical actual pose and the change in the current pose, so as to obtain the trajectory prediction pose of the vehicle to be located at the current moment.

[0115] The underground vehicle positioning device provided in this embodiment of the invention can execute the underground vehicle positioning method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.

[0116] Example 4

[0117] Figure 4 A schematic diagram of an electronic device 10 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0118] like Figure 4 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 may also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0119] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0120] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as downhole vehicle positioning methods.

[0121] In some embodiments, the downhole vehicle positioning method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or mounted on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the downhole vehicle positioning method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the downhole vehicle positioning method by any other suitable means (e.g., by means of firmware).

[0122] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0123] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0124] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0125] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0126] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), target blockchain networks, and the Internet.

[0127] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0128] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0129] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for locating underground vehicles, characterized in that, include: Based on the set of historical candidate poses of the vehicle to be located at the previous time corresponding to the current time and the predetermined changes in the historical poses of the vehicle to be located at multiple historical times, the set of current candidate poses of the vehicle to be located at the current time is determined. The semantic features of the roadway corresponding to the vehicle to be located at the current time are obtained, and semantic matching is performed based on the semantic features of the roadway, the semantic map corresponding to the mine roadway where the vehicle to be located is located, and the current candidate pose set to determine the semantic matching pose of the vehicle to be located at the current time; wherein, the semantic features of the roadway are used to indicate the roadway traffic objects existing on the roadway; Point cloud matching is performed based on the semantic matching pose, the scanned point cloud data of the vehicle to be located at the current time, and the point cloud map corresponding to the mine roadway to determine the point cloud matching pose of the vehicle to be located at the current time. The point cloud matching pose and the pre-determined trajectory prediction pose of the vehicle to be located at the current time are fused to obtain the target positioning pose of the vehicle to be located at the current time. The step of performing semantic matching based on the roadway semantic features, the semantic map corresponding to the mine roadway where the vehicle to be located is located, and the current candidate pose set to determine the semantic matching pose corresponding to the vehicle to be located at the current time includes: for each current candidate pose in the current candidate pose set, determining a target weight corresponding to the current candidate pose based on the current candidate pose, the roadway semantic features, and the semantic map; and determining the semantic matching pose corresponding to the vehicle to be located at the current time from the multiple current candidate poses based on the target weights corresponding to the multiple current candidate poses. The step of performing point cloud matching based on the semantic matching pose, the scanned point cloud data corresponding to the vehicle to be located at the current time, and the point cloud map corresponding to the mine roadway to determine the point cloud matching pose corresponding to the vehicle to be located includes: determining the product between the semantic matching pose and a pre-determined transformation parameter matrix to obtain the point cloud pose to be processed; processing the point cloud pose to be processed, the scanned point cloud data, and the actual point cloud data corresponding to the scanned point cloud data on the point cloud map according to a preset point cloud matching algorithm to obtain a matching error; adjusting the point cloud pose to be processed with the minimum matching error as the optimization target, and taking the point cloud pose to be processed corresponding to the minimum matching error as the point cloud matching pose corresponding to the vehicle to be located.

2. The underground vehicle positioning method according to claim 1, characterized in that, The historical candidate pose set includes multiple historical candidate poses corresponding to the vehicle to be located at the previous time; determining the current candidate pose set corresponding to the vehicle to be located at the current time based on the historical candidate pose set corresponding to the vehicle to be located at the previous time and the pre-determined historical pose change amounts corresponding to the vehicle to be located at multiple historical time points includes: Obtain the set of historical candidate poses of the vehicle to be located at the current time corresponding to the previous time and the changes in historical poses at multiple historical times; For each historical candidate pose in the historical candidate pose set, determine the product between the historical candidate pose and multiple historical pose changes to obtain the current candidate pose corresponding to the current moment; The set of current candidate poses corresponding to the current time of the vehicle to be located is determined based on the multiple current candidate poses corresponding to the current time.

3. The underground vehicle positioning method according to claim 1, characterized in that, The step of obtaining the alleyway semantic features corresponding to the vehicle to be located at the current time includes: By pre-setting a camera on the vehicle to be located, images of the mine roadway where the vehicle is located are acquired, so as to obtain the roadway image corresponding to the vehicle at the current time. Semantic features are extracted from the tunnel image to obtain the tunnel semantic features corresponding to the vehicle to be located at the current time.

4. The underground vehicle positioning method according to claim 1, characterized in that, The step of determining the target weight corresponding to the current candidate pose based on the current candidate pose, the alleyway semantic features, and the semantic map includes: Determine the product between the current candidate pose and the alleyway semantic features to obtain the visual detection features corresponding to the current candidate pose; Determine the distance between the visual detection features and the semantic feature elements in the semantic map that correspond to the alleyway semantic features, so as to obtain the semantic matching distance corresponding to the current candidate pose; The target weight corresponding to the current candidate pose is determined based on the semantic matching distance.

5. The underground vehicle positioning method according to claim 1, characterized in that, The step of determining the semantic matching pose corresponding to the vehicle to be located from the multiple current candidate poses based on the target weights corresponding to the multiple current candidate poses includes: The target weight with the largest weight value is determined based on multiple target weights, and the current candidate pose corresponding to the determined target weight is used as the semantic matching pose corresponding to the vehicle to be located.

6. The method for positioning underground vehicles according to claim 1, characterized in that, The step of fusing the point cloud matching pose with the pre-determined trajectory prediction pose of the vehicle to be located at the current time to obtain the target positioning pose of the vehicle to be located at the current time includes: The point cloud matching pose and the trajectory prediction pose are interpolated and fused according to the fusion localization algorithm, and the fused pose is used as the target localization pose of the vehicle to be located at the current time.

7. The underground vehicle positioning method according to claim 1, characterized in that, Also includes: Obtain the historical actual pose of the vehicle to be located at the previous moment and the change in the current pose at the current moment; The product between the historical actual pose and the change in the current pose is determined to obtain the trajectory prediction pose of the vehicle to be located at the current moment.

8. A positioning device for underground vehicles, characterized in that, include: The candidate pose set determination module is used to determine the current candidate pose set of the vehicle to be located at the current time based on the historical candidate pose set corresponding to the previous time of the vehicle to be located at the current time and the predetermined historical pose change amount of the vehicle to be located at multiple historical times. The semantic matching pose determination module is used to obtain the roadway semantic features corresponding to the vehicle to be located at the current time, and to perform semantic matching based on the roadway semantic features, the semantic map corresponding to the mine roadway where the vehicle to be located is located, and the current candidate pose set to determine the semantic matching pose corresponding to the vehicle to be located at the current time; wherein, the roadway semantic features are used to indicate the roadway traffic objects existing on the roadway; The point cloud matching pose determination module is used to perform point cloud matching based on the semantic matching pose, the scanned point cloud data of the vehicle to be located at the current time, and the point cloud map corresponding to the mine roadway, so as to determine the point cloud matching pose of the vehicle to be located at the current time. The positioning pose determination module is used to perform pose fusion between the point cloud matching pose and the pre-determined trajectory prediction pose of the vehicle to be located at the current time, so as to obtain the target positioning pose of the vehicle to be located at the current time. The semantic matching pose determination module includes a pose weight determination unit and a semantic matching pose determination unit; wherein, the pose weight determination unit is used to determine a target weight corresponding to each current candidate pose in the current candidate pose set, based on the current candidate pose, the alleyway semantic features, and the semantic map; the semantic matching pose determination unit is used to determine the semantic matching pose of the vehicle to be located at the current time from the multiple current candidate poses based on the target weights corresponding to the multiple current candidate poses; The point cloud matching pose determination module includes: a point cloud pose determination unit, a matching error determination unit, and a point cloud matching pose determination unit; wherein, the point cloud pose determination unit is used to determine the product between the semantic matching pose and a pre-determined transformation parameter matrix to obtain the point cloud pose to be processed; the matching error determination unit is used to process the point cloud pose to be processed, the scanned point cloud data, and the actual point cloud data on the point cloud map corresponding to the scanned point cloud data according to a preset point cloud matching algorithm to obtain the matching error; the point cloud matching pose determination unit is used to adjust the point cloud pose to be processed with the minimum matching error as the optimization target, and take the point cloud pose to be processed corresponding to the minimum matching error as the point cloud matching pose corresponding to the vehicle to be located.

Images