Robot moving environment constraint construction method, robot, and storage medium

Abstract

The application provides a robot moving environment constraint construction method, a computer program product, a robot and a storage medium, and the robot moving environment constraint construction method comprises the following steps: generating geometric representation information of a channel structure boundary in a robot traveling environment based on environment perception data; and defining the geometric representation information of the channel structure boundary as an impassable area.

Description

Technical Field

[0001] This application relates to the field of robot control technology, and in particular to a method for constructing constraints for robot movement environment, a computer program product, a robot, and a storage medium. Background Technology

[0002] In applications such as poultry house aisles, underground utility tunnels, and mines, these aisles are typically characterized by their narrowness, repetitive structure, strong environmental constraints, and limited traversable areas. In such environments, mobile devices need to remain within the aisle for extended periods during operation and must cope with potential obstructions, narrowing, bends, and dynamic obstacles.

[0003] In existing technologies, it is usually necessary to modify the site to construct guidance methods such as tracks and magnetic strips, which is costly. Alternatively, movement control is based on strong prior assumptions about the scene, resulting in insufficient robustness in complex or changing channel environments, making it difficult to achieve stable and continuous autonomous movement. Summary of the Invention

[0004] To address the existing technical problems, this application provides a method for constructing robot mobile environment constraints, a computer program product, a robot, and a storage medium that can improve the reliability of robot autonomous movement.

[0005] In a first aspect, embodiments of this application provide a method for constructing constraints in a robot's mobile environment, comprising: generating geometric representation information of the boundary of a channel structure in the robot's travel environment based on environmental perception data; and defining the geometric representation information of the boundary of the channel structure as a non-passable area.

[0006] In a second aspect, a computer program product is provided, comprising a computer program that, when executed by a processor, implements the robot mobility environment constraint construction method as described in any embodiment of this application.

[0007] Thirdly, a robot is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the robot mobility environment constraint construction method as described in any embodiment of this application.

[0008] Fourthly, a storage medium is provided, on which a computer program is stored, which, when executed by a processor, implements the robot mobile environment constraint construction method as described in any embodiment of this application.

[0009] In the robot mobility environment constraint construction method provided in the above embodiments, geometric representation information of the channel structure boundary in the robot's travel environment is generated based on environmental perception data; the geometric representation information of the channel structure boundary is defined as an impassable area. Thus, the geometric representation information of the channel structure boundary can be quickly and accurately determined using environmental perception data, thereby accurately identifying the channel structure boundary as an impassable area. This facilitates the rapid and accurate generation of a driving trajectory that automatically avoids the channel structure boundary. It eliminates the need for prior modification of the site to form guide devices such as tracks, and also eliminates the need for trajectory planning based on prior scene knowledge, reducing reliance on human intervention. By fully perceiving and utilizing the characteristics of the channel structure boundary, the reliability and flexibility of the robot's autonomous movement in diverse and complex channel scenarios are improved.

[0010] The computer program product, robot, and storage medium provided in the above embodiments belong to the same concept as the corresponding robot mobile environment constraint construction method embodiments, and thus have the same technical effects as the corresponding robot mobile environment constraint construction method embodiments, which will not be repeated here. Attached Figure Description

[0011] Figure 1 This is a flowchart illustrating a method for constructing constraints for a robot's mobile environment in one embodiment of this application. Figure 2 This is a schematic diagram of the channel structure and the channel structure boundary in one embodiment of this application; Figure 3 This is a schematic diagram of an obstacle area in one embodiment of this application; Figure 4 This is a schematic diagram of the driving trajectory in one embodiment of this application; Figure 5 This is a flowchart illustrating a method for constructing constraints for a robot's mobile environment in another embodiment of this application. Figure 6 This is a schematic diagram of a robot mobile control system according to one embodiment of this application; Figure 7 This is a schematic diagram of the robot structure in one embodiment of this application. Detailed Implementation

[0012] The technical solution of this application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0013] To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings. The described embodiments should not be regarded as limitations on this application. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0014] In the following description, the phrase "some embodiments" refers to a subset of all possible embodiments. It should be noted that "some embodiments" can be the same subset or different subsets of all possible embodiments, and can be combined with each other without conflict.

[0015] In the following description, the terms "first," "second," and "third" are used merely to distinguish similar objects and do not represent a specific ordering of the objects. It is understood that "first," "second," and "third" can be interchanged in a specific order or sequence where permitted, so that the embodiments of this application described herein can be implemented in an order other than that illustrated or described herein. It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other.

[0016] In related technologies, mobile devices such as robots typically achieve movement in the above scenarios using the following methods: 1. Manual remote control or guidance methods such as tracks and magnetic strips. These methods require secondary modifications to the site, which are costly and not feasible in some scenarios. 2. Positioning and navigation methods based on high-precision maps require the scene to be mapped in advance, but mapping is difficult in narrow passage environments, and positioning errors tend to accumulate with driving distance. 3. Localization or relative pose estimation methods based on environmental feature extraction: These methods are highly dependent on environmental features, and the features of different sites vary greatly, often requiring manual adjustment or calibration for specific sites.

[0017] When the passage is obstructed, narrowed, curved, or the environment changes, existing technologies usually have the following problems: the track or magnetic strip method is difficult to cope with changes in temporary obstacles; in the high-precision map method, the geometric constraints of the passage are difficult to be directly reflected in the positioning results; and the feature extraction method is prone to failure when the environmental features change.

[0018] The methods described above typically only output positioning results or simple obstacle avoidance information, failing to explicitly reflect the structural constraints of the passageway, thus making it difficult to achieve stable and continuous motion control in response to changes in the passageway structure. Furthermore, existing technologies generally rely on high-precision maps, manual calibration, or human intervention, further limiting their applicability in complex passageway environments.

[0019] Based on this, such as Figure 1 As shown, one embodiment of this application provides a method for constructing constraints for a robot's mobile environment, including: S10: Generate geometric representation information of the channel structure boundary in the robot's travel environment based on environmental perception data.

[0020] In one embodiment, environmental perception data is data obtained by detecting the environment in front of the robot. For example, environmental perception data may include image data and / or point cloud data, where image data is environmental image data representing the path to be traveled, and point cloud data may be three-dimensional point cloud data, etc.

[0021] In one embodiment, a passage is a space in the travel environment that the robot can traverse. The passage structure is a boundary or constraint structure that encloses the passage. The boundary of the passage structure may include, for example, one or both boundaries of the left and right sides of the passage, as well as the upper and / or lower boundaries of the passage.

[0022] In one embodiment, the geometric representation information can be linear geometric elements, such as at least one element including straight line segments, curve segments, and endpoints. The geometric representation information can characterize the specific location of the channel structure boundary and its relative positional relationship with other areas in the travel environment.

[0023] In one embodiment, generating geometric representation information of the channel structure boundary in the robot's travel environment based on environmental perception data may include: abstracting the channel structure boundary into a continuous curve form of the channel structure boundary line by fitting the point cloud set corresponding to the channel in the point cloud data. The continuous curve form of the channel structure boundary line is the aforementioned geometric representation information.

[0024] S20: Define the geometric representation information of the channel structure boundary as an impassable region.

[0025] In one embodiment, an impassable region represents an area in the travel environment that is not allowed to be traversed and needs to be avoided. For example, an impassable region may include geometric representation information of the boundary of the aforementioned passage structure, and may also include geometric representation information corresponding to obstacle areas, etc.

[0026] In one embodiment, impassable areas represent environmental occupancy information within the travel environment. For example, the portion defined as an impassable area can be represented in a grid format, such as... Figure 2 As shown.

[0027] In one embodiment, after step S20, the method may further include: generating a driving trajectory through the passage based on the impassable area.

[0028] In one embodiment, generating a driving trajectory through the passage based on the impassable area may include: generating a driving trajectory through the passage based on the impassable area and odometer information. The odometer information may be determined based on image data and point cloud data.

[0029] In one embodiment, the travel trajectory can be correlated with the geometric representation information of the channel structure boundary. For example, the distance between the travel trajectory and any segment of the channel structure boundary represented by the geometric representation information is greater than a preset distance. The preset distance can be determined based on the geometric representation information of the channel structure boundary and the robot's size information.

[0030] For example, the travel trajectory can be determined based on the centerlines of the two sides of the passage structure represented by geometrical representation information and obstacle regions in the impassable areas. The obstacle regions characterize objects that may affect the robot's movement. For instance, the travel trajectory may partially or completely coincide with the centerlines of the two sides of the passage structure.

[0031] In this way, by utilizing environmental perception data, the geometric representation information of the channel structure boundary can be quickly and accurately determined, thereby accurately identifying the channel structure boundary as a non-passable area. This facilitates the rapid and accurate generation of driving trajectories that automatically avoid the channel structure boundary. It eliminates the need for pre-modification of the site to form guide devices such as tracks, and also eliminates the need for trajectory planning based on prior knowledge of the scene, reducing the reliance on human intervention. By fully perceiving and utilizing the characteristics of the channel structure boundary, the reliability and flexibility of the robot's autonomous movement in diverse and complex channel scenarios are improved.

[0032] In some embodiments, generating geometric representation information of the channel structure boundary in the robot's travel environment based on environmental perception data may include: Based on preset categories, semantic segmentation is performed on environmental perception data to obtain classification results; Based on the classification results, the channel structure is determined, and the spatial information corresponding to the channel structure is obtained; Geometric representation information of the channel structure boundary is generated based on the spatial information.

[0033] In one embodiment, a preset category refers to the category to which different areas in the travel environment belong. For example, a preset category may include, but is not limited to, at least one of the following: ground area, passage structure and / or obstacle area.

[0034] In one embodiment, semantic segmentation of environmental perception data based on a preset category to obtain a classification result may include: performing semantic segmentation of the environmental perception data based on a preset category to obtain a segmentation result; and constraining or correcting the segmentation result based on the corresponding depth data in the image data to obtain a classification result. For example, the image data may be depth image data acquired by a depth camera.

[0035] In one embodiment, obtaining the classification result may include: obtaining the classification result and performing temporal consistency processing on the classification result. This can reduce the fluctuation of the segmentation result caused by perceptual noise or environmental changes, and improve the stability of channel structure recognition.

[0036] In one embodiment, the semantic segmentation output classification result is a pixel-level classification result. Optionally, after obtaining the classification result, a corresponding region mask can be generated based on the classification result to generate the geometric representation information of the channel structure boundary with the spatial information.

[0037] In one embodiment, the spatial information corresponding to the channel structure can be determined based on the point cloud data corresponding to the channel structure.

[0038] In this way, different regions in the environmental image data are classified according to preset categories, so that the channel structure can be accurately determined. Then, based on the corresponding spatial information, the geometric representation of all boundaries of the channel structure can be completely obtained, which is conducive to forming comprehensive and accurate spatial geometric representation information and improving the accuracy of trajectory planning.

[0039] In some embodiments, the semantic segmentation of environmental perception data based on a preset category may include: Acquire environmental perception data corresponding to the robot's travel environment, including image data and point cloud data; Registration is performed based on the point cloud data and the image data; The registered image data is semantically segmented based on a preset category.

[0040] In one embodiment, image data can be acquired using a depth camera included in the robot; for example, the image data may be depth image data. Point cloud data can be acquired using a 3D radar module included in the robot, or it can also be acquired using a depth camera included in the robot.

[0041] For example, a depth camera can be an integrated device that simultaneously outputs depth image data and point cloud data, and there is a one-to-one correspondence between the image pixels and the points in the point cloud data.

[0042] In one embodiment, registration based on the point cloud data and the image data may include: obtaining a pre-stored registration relationship; and registering the point cloud data and the image data based on the registration relationship.

[0043] In one embodiment, obtaining the spatial information corresponding to the channel may include: filtering out a set of point clouds corresponding to the channel structure from the point cloud data, wherein the set of point clouds represents the spatial information corresponding to the channel structure.

[0044] In one embodiment, the geometric representation information generated based on spatial information is the geometric representation information of the channel structure boundary. The channel structure boundary may include the boundary of at least one side of the channel structure, such as the left or right boundary, or it may include the boundaries of both the left and right sides.

[0045] In this way, by registering the collected depth images and 3D point cloud data, the position of each type of area in the forward travel environment can be accurately represented. Then, the point cloud set corresponding to the channel structure can be selected from the point cloud data, which can be used to accurately determine the corresponding spatial geometric representation and improve the accuracy of channel recognition.

[0046] In some embodiments, generating a travel trajectory through the passage based on the impassable area may include: Odometry information is determined based on the registered image data and the point cloud data; Based on the odometer information and the impassable area, a driving trajectory is generated to pass through the passage.

[0047] In one embodiment, odometry information can be collected based on devices such as odometry included in the robot (e.g., wheeled odometry) and used to characterize the robot's pose, distance traveled, and other positioning status.

[0048] In one embodiment, odometer information can be determined from image data and point cloud data, for example, based on the depth data corresponding to points in the point cloud data in the image data.

[0049] In one embodiment, before generating a driving trajectory through the passage based on the odometer information and the impassable area, the method may further include: determining whether there is a dynamic obstacle; and if there is a dynamic obstacle, obtaining a prediction result of the motion state of the dynamic obstacle.

[0050] For example, dynamic obstacles may include obstacles such as poultry or animals.

[0051] In one embodiment, the method may further include: supplementing and updating impassable areas based on the predicted motion state of dynamic obstacles.

[0052] In one embodiment, after generating a driving trajectory through the passage based on the odometer information and the impassable area, the method may further include: correcting the driving trajectory based on the motion state prediction results of the dynamic obstacles.

[0053] In this way, by using the positional relationships and depth data represented in image data and point cloud data, the odometry information of the robot's current motion and position can be determined. Then, by combining the odometry information, a driving trajectory that is more accurately matched with the robot's current pose and positioning can be generated, thereby improving the accuracy of trajectory planning.

[0054] In some embodiments, the preset categories include: ground area, channel structure, and obstacle area; the channel structure is a boundary structure that encloses and forms the channel, the ground area is a drivable area within the channel, and the obstacle area is located in the ground area.

[0055] The step of defining the geometric representation information of the channel structure boundary as a non-passable area may include: defining the identified obstacle area and the geometric representation information of the channel structure boundary as non-passable areas.

[0056] In one embodiment, the obstacle region may include a fixed obstacle region and a dynamic obstacle region. The relationship between the obstacle region and the passage structure can be as follows: Figure 3 As shown.

[0057] In one embodiment, when the obstacle region includes a dynamic obstacle region, after generating a driving trajectory through the passage based on the impassable area, the method may further include: continuing to acquire environmental perception data to determine whether the driving trajectory is blocked by the dynamic obstacle region; if so, updating the driving trajectory based on the dynamic obstacle region.

[0058] For example, updating the driving trajectory based on dynamic obstacle regions can include generating a driving trajectory that avoids the dynamic obstacle regions and the boundaries of the passage structure. For example, generating a trajectory such as... Figure 4 The driving trajectory shown.

[0059] In this way, by defining the obstacle area and the boundary of the passage structure together as an impassable area, the obstacle avoidance requirement of the passage structure boundary can be met. No prior knowledge or track setting is required, and a driving trajectory that avoids obstacles and travels in the passage can be directly generated.

[0060] In some embodiments, generating a travel trajectory through the passage based on the impassable area includes: Based on the geometric representation information of the channel structure boundary, the centerline of the channel is determined; A driving trajectory through the passage is generated based on the centerline and the impassable area.

[0061] In one embodiment, the centerline of the channel is the centerline between the two boundaries of the channel. For example, any point on the centerline is equidistant from the two boundaries of the channel.

[0062] In one embodiment, generating a driving trajectory through the passage based on the centerline and the impassable area may include: generating a driving trajectory through the passage based on the centerline and the obstacle area. For example, when the impassable area does not contain obstacle areas, the driving trajectory may completely coincide with the centerline; when the impassable area contains obstacle areas, the driving trajectory may partially coincide with the centerline, and the non-coincident portion is a trajectory that bypasses the obstacle areas, for example... Figure 4 As shown.

[0063] In this way, generating the driving trajectory based on the centerline can minimize the risk of collision between the driving trajectory and the boundary of the channel structure, thus improving the reliability of trajectory planning.

[0064] In some embodiments, such as when the geometric representation information of the channel structure boundary corresponds only to one side boundary of the channel, generating a driving trajectory through the channel based on the impassable area may include: Obtain the robot's size information; Set the offset distance based on the size information; A driving trajectory through the passage is generated based on the offset distance and the impassable area.

[0065] In one embodiment, setting the offset distance based on the size information may include setting the offset distance based on the channel structure boundary corresponding to the size information and the geometric representation information.

[0066] In one embodiment, the geometric representation information of the channel structure boundary corresponds only to one side boundary of the channel, meaning that the generated geometric representation information only includes the geometric representation information of one side boundary of the channel. For example, it may only include the geometric representation information of the left or right boundary.

[0067] In one embodiment, the robot's size information may include at least the robot's width data. Here, the width data refers to the width in the direction perpendicular to the direction of travel (i.e., the forward direction), that is, the width in the direction from one boundary of the channel to the other.

[0068] In one embodiment, the offset distance is the distance that needs to be maintained between the robot's center position and the boundary of the channel structure. For example, it refers to the distance that needs to be maintained between the robot's center position and the boundary of one side of the channel corresponding to the geometric representation information.

[0069] Thus, when only the geometric representation of one side boundary of the channel can be obtained, the centerline of the channel cannot be accurately determined. At this time, by using the offset distance set between the robot size and the currently known channel structure boundary, the robot can be precisely controlled to travel along a trajectory that maintains a certain distance from the boundary, avoiding collisions with the channel structure boundary and improving the reliability of trajectory planning.

[0070] As one possible implementation, embodiments of this application provide a method for controlling intra-column movement of a robot, such as... Figure 5 As shown, it can be specifically applied to, for example... Figure 6 The core idea of ​​the system shown is to achieve stable and autonomous movement of the robot in a narrow channel by explicitly perceiving and modeling the channel structure and using the channel structure as an important constraint for motion planning and control, without relying on global positioning or high-precision maps.

[0071] Specifically, using images and point cloud data acquired by a depth camera, semantic segmentation of the environment is performed, dividing the scene into different categories including at least the ground, passage structures, and obstacles; based on this, spatial geometric elements corresponding to the passage structures are extracted from the segmentation results to construct an environment representation for navigation; combined with local odometry information, a motion trajectory that satisfies the constraints of the passage structures is generated, and the robot is controlled to travel along the trajectory.

[0072] For example, image data and point cloud data of the environment are acquired using a depth camera mounted on the robot as environmental perception data, and the image data and point cloud data are registered. A semantic segmentation module is used to perform semantic segmentation processing on the image data, dividing the environment into at least the following categories: ground area, passage structure, and obstacle area. The passage structure is used to characterize the boundary or constraint structure of the passage, and the obstacle area is used to characterize objects that may affect the robot's movement.

[0073] The channel structure construction module is used to extract spatial information corresponding to the channel structure from point cloud data based on semantic segmentation results, and abstract the channel structure into at least one set of linear geometric elements to represent the channel boundary. In one implementation, the channel structure can be represented as at least one side of the left and right boundaries of the channel; when only one boundary is detected, a fixed offset distance is preset according to the robot's own size to determine the feasible driving area within the channel. In another implementation, a centerline representing the middle of the channel can be further generated to guide the robot to travel along the middle of the channel.

[0074] The environment occupancy information construction module is used to construct environment occupancy information for navigation based on point cloud data and semantic segmentation results. The area corresponding to the channel structure is set as an impassable area, which has the same impassable attribute as the obstacle area.

[0075] The trajectory generation and planning module combines the geometric representation of the channel structure, environmental occupancy information, and local odometry information to generate a motion trajectory that satisfies the channel structure constraints. When no obstacles enter the predicted travel path, the robot generates a basic trajectory along the center region of the channel and remains within the channel structure constraints. When an obstacle is detected entering the predicted travel path, the robot generates an obstacle avoidance trajectory to avoid the obstacle while still satisfying the channel structure constraints.

[0076] The motion control module is used to control the robot's motion based on the generated motion trajectory and local odometry information, and to control the robot's actuators to make the robot travel along the trajectory within the channel.

[0077] As one implementation method, the above method may include the following steps: S1. Environmental perception and data acquisition.

[0078] In this embodiment, a depth camera is installed on the robot. The depth camera is configured with a combined forward and downward view, and the lower edge of its field of view is tangent to the robot's own boundary, thereby covering the entire ground area in the robot's forward direction and ensuring that the passage structure is within the perception field of view.

[0079] In one embodiment, the depth camera is an integrated device that can simultaneously output depth image data and point cloud data, with a one-to-one correspondence between image pixels and point cloud data points. The registration relationship between the image data and point cloud data is completed internally by the device, requiring no additional online calibration or external correction.

[0080] In one embodiment, during robot operation, the depth camera continuously acquires environmental data at a real-time frequency. When missing, abnormal, or unavailable sensing data is detected, the system triggers a safety mechanism, causing the robot to stop moving and reporting the status to the upper-level system.

[0081] S2. Semantic segmentation of channel structure and obstacles.

[0082] After acquiring environmental perception data, semantic segmentation is performed on the image data (i.e., environmental image data). Semantic segmentation uses image data as the primary input and combines it with corresponding depth data to constrain or correct the segmentation results. The output of semantic segmentation is a pixel-level classification result, which includes at least the following categories: ground region; channel structure region; obstacle region. When needed, the pixel-level classification result can be further used to generate corresponding region masks for subsequent geometric processing or spatial analysis.

[0083] In one embodiment, the channel structure is considered a necessary perceptual element for the robot's normal operation. When a channel structure region is detected to be unrecognizable or continuously missing, the system ceases to perform movement control within the channel and instead delegates control to a higher-level system. In an alternative implementation, to reduce fluctuations in segmentation results caused by perceptual noise or environmental changes, temporal consistency processing can be applied to the semantic segmentation results to improve the stability of channel structure recognition.

[0084] S3, Geometric representation of the channel structure.

[0085] Based on semantic segmentation results, a point cloud set corresponding to the channel structure region is selected from the point cloud data, and a spatial geometric representation of the channel structure is generated based on the point cloud set. In this embodiment, the channel structure is abstracted into a continuous curve-shaped channel structure boundary line by fitting the point cloud set. The channel structure boundary line can represent at least one of the left and right boundaries of the channel. When both left and right boundaries of the channel are detected simultaneously, the channel centerline can be calculated based on the positional relationship between the two boundaries. The centerline is used to describe the driving guidance path in the middle of the channel. When only one side of the channel structure boundary is detected, a feasible driving area is generated inside the single-sided boundary based on a preset configurable offset distance according to the robot's own width, which is used to replace the channel centerline. The geometric representation of the channel structure is updated periodically, and the update frequency can be configured according to the performance of the computing unit.

[0086] S4, Environmental Occupation Information Construction.

[0087] Environmental occupancy information for navigation is constructed based on point cloud data and semantic segmentation results. This information is represented using a local occupancy grid and updated in real-time with reference to the robot's current pose. Within this occupancy grid: areas corresponding to passage structures are set as permanently impassable areas; areas corresponding to obstacles are set as impassable areas. For dynamic obstacles, their occupancy status can be updated with a lag to reduce the impact of instantaneous perception fluctuations on the planning results. This environmental occupancy information not only describes currently traversable areas but can also be used in subsequent trajectory prediction and evaluation processes.

[0088] S5. Trajectory Generation and Planning.

[0089] After obtaining the geometric representation of the channel structure, environmental occupancy information, and odometry information, the system executes the trajectory generation and planning process. The system uses the currently planned trajectory as the predicted travel path range to determine whether obstacles might affect the robot's current travel path. If no obstacles are detected within the predicted travel path range (i.e., the travel trajectory), the system generates a basic trajectory along the channel centerline, keeping the robot within the channel structure constraints. When an obstacle is detected entering the predicted travel path range, the system adjusts the trajectory generation process based on environmental occupancy information and the predicted motion state of dynamic obstacles, generating an obstacle avoidance trajectory while satisfying the channel structure's non-crossing constraints. In the absence of obstacles, a basic trajectory along the channel centerline is generated; when an obstacle is detected entering the predicted travel path range, an obstacle avoidance trajectory that satisfies the channel structure constraints is generated.

[0090] In the presence of dynamic obstacles, the predicted motion state of these obstacles can be used to update environmental occupancy information or the trajectory evaluation process, thereby affecting the feasibility judgment or cost assessment of the trajectory. When trajectory planning fails, the system controls the robot to stop moving and reports the planning failure status to the upper-level system; after the status is updated, the trajectory generation process can be re-executed.

[0091] S6, Motion Control.

[0092] Based on the generated motion trajectory, the system outputs control commands in the form of path curves to the robot's motion control module. The motion control process is executed based on local odometry information obtained through the fusion of vision and inertial navigation, enabling the robot to travel within the planned trajectory. When the system detects that the robot deviates from the channel's structural constraints, it corrects the motion trajectory and reduces the robot's speed to improve safety. The motion control is executed in real-time. When a control anomaly or execution failure is detected, the system triggers a safety mechanism, stopping the robot and notifying the upper-level system of the relevant status.

[0093] In one optional implementation, the system can adaptively adjust the geometric representation of the channel structure and the trajectory generation process to adapt to changes in channel width, thus accommodating environments with varying channel widths. In another optional implementation, targets with obvious dynamic characteristics, such as poultry and animals, can be identified as dynamic obstacles through continuous displacement detection and included in the trajectory generation process. In a further implementation, the system can be equipped with an anomaly detection mechanism; when a perception anomaly, planning anomaly, or control anomaly is detected, a safe stop is executed and the upper-level system is notified.

[0094] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the robot mobile environment constraint construction method as described in any embodiment of this application.

[0095] This application also provides a robot, such as an inspection robot. Figure 7 This is a structural diagram of a robot according to an embodiment of this application, such as... Figure 7 As shown, the robot includes a processor 41, a communication interface 42, a memory 43, and a communication bus 44. The processor 41, the communication interface 42, and the memory 43 communicate with each other through the communication bus 44. The memory 43 is used to store computer programs. When the processor 41 executes the program stored in the memory 43, it implements the steps of the method described in any one or more of the aforementioned method embodiments.

[0096] The communication bus mentioned above can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used to represent it in the diagram, but this does not mean that there is only one bus or one type of bus.

[0097] The communication interface is used for communication between the aforementioned terminal and other devices.

[0098] The memory may include random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0099] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0100] In another embodiment provided in this application, a computer-readable storage medium is also provided, which stores instructions that, when executed on a computer, cause the computer to perform the method described in any of the above embodiments.

[0101] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in the embodiments of this application are generated entirely or partially. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid state disk (SSD)).

[0102] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0103] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for constructing environmental constraints for robot movement, characterized in that, include: Generate geometric representation information of the channel structure boundary in the robot's travel environment based on environmental perception data; The geometric representation information of the boundary of the channel structure is defined as an impassable region.

2. The method for constructing constraints for robot movement environment according to claim 1, characterized in that, Also includes: A driving trajectory through the passage is generated based on the impassable area.

3. The method for constructing constraints for robot movement environment according to claim 2, characterized in that, The generation of geometric representation information of the channel structure boundary in the robot's travel environment based on environmental perception data includes: Based on preset categories, semantic segmentation is performed on environmental perception data to obtain classification results; Based on the classification results, the channel structure is determined, and the spatial information corresponding to the channel structure is obtained; Geometric representation information of the channel structure boundary is generated based on the spatial information.

4. The method for constructing constraints for robot movement environment according to claim 3, characterized in that, The semantic segmentation of environmental perception data based on preset categories includes: Acquire environmental perception data corresponding to the robot's travel environment, including image data and point cloud data; Registration is performed based on the point cloud data and the image data; The registered image data is semantically segmented based on a preset category; The step of obtaining the spatial information corresponding to the channel structure includes: The point cloud set corresponding to the channel structure is selected from the point cloud data, and the point cloud set represents the spatial information corresponding to the channel structure.

5. The method for constructing constraints for robot movement environment according to claim 4, characterized in that, The generation of a driving trajectory through the passage based on the impassable area includes: Odometry information is determined based on the registered image data and the point cloud data; Based on the odometer information and the impassable area, a driving trajectory is generated to pass through the passage.

6. The method for constructing constraints for robot movement environment according to claim 3, characterized in that, The preset categories include: ground area, passage structure, and obstacle area; the passage structure is the boundary structure that encloses and forms the passage, the ground area is the drivable area within the passage, and the obstacle area is located within the ground area; The step of defining the geometric representation information of the channel structure boundary as a non-passable region includes: The geometric representation information of the identified obstacle area and the boundary of the channel structure is defined as an impassable area.

7. The method for constructing constraints for robot movement environment according to claim 2, characterized in that, The generation of a driving trajectory through the passage based on the impassable area includes: Based on the geometric representation information of the channel structure boundary, the centerline of the channel is determined; A driving trajectory through the passage is generated based on the centerline and the impassable area.

8. The method for constructing constraints for robot movement environment according to claim 2, characterized in that, The generation of a driving trajectory through the passage based on the impassable area includes: Obtain the robot's size information; Set the offset distance based on the size information; A driving trajectory through the passage is generated based on the offset distance and the impassable area.

9. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the robot movement environment constraint construction method as described in any one of claims 1 to 8.

10. A robot, characterized in that, The robot includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When executed by the processor, the computer program implements the robot mobility environment constraint construction method as described in any one of claims 1 to 8.

11. A storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the robot movement environment constraint construction method as described in any one of claims 1 to 8.

Images