A 3D environment overlay method for online sub-region updating and hierarchical planning

Abstract

This invention discloses a 3D environment coverage method based on online sub-region updating and hierarchical planning. Through online sub-region updating and hierarchical planning, it significantly improves exploration efficiency and reduces computational cost. First, a fast environment preprocessing method is used, examining only updated boundaries to extract frontier points, thereby reducing computational burden. Second, a sparse topology graph is constructed, discretizing the space and using a KD-tree to store topology nodes, avoiding time-consuming path searches on the 3D voxel map. Furthermore, a hierarchical sub-region structure is proposed, dividing the environment into multiple adaptive sub-regions and updating environment information in real time to support subsequent exploration planning. A hierarchical planning strategy is adopted, modeling environment coverage as a traveling salesman problem to solve the global coverage path, considering the overall path length for sub-region visits. Compared to greedy strategies, this method provides a long-term perspective, improving exploration efficiency. Finally, the method's efficiency and feasibility are verified through simulation and field tests.

Description

Technical Field

[0001] This invention belongs to the field of ground robot technology, specifically relating to a three-dimensional environment coverage method for online sub-region updating and hierarchical planning. Background Technology

[0002] In recent years, the widespread application of ground mobile robots in key areas such as search and rescue, environmental monitoring, and industrial automation has significantly improved operational efficiency and ensured personnel safety under extreme or dangerous conditions. Their core advantage lies in their ability to autonomously navigate and explore unknown environments without direct human intervention. Despite decades of research and development, resulting in various autonomous exploration technologies and methods, achieving efficient and comprehensive autonomous exploration in large-scale environments with complex topologies remains a significant challenge.

[0003] Mainstream exploration frameworks primarily include boundary-based methods and sampling-based methods. Boundary-based methods focus on identifying and prioritizing the exploration of the boundary zone between free space and unknown areas, while sampling methods guide the exploration process by randomly selecting and evaluating potential viewpoints. Both methods gradually cover the entire environment through iterative searching and evaluation. However, some limitations still hinder the development of autonomous exploration, as described below:

[0004] First, there is the high computational cost. Accurate boundary detection and potential viewpoint benefit assessment involve complex voxel occupancy checks and ray tracing algorithms, resulting in enormous computational resource consumption. For resource-constrained edge computing platforms, meeting real-time requirements is particularly difficult. Second, there is inefficiency. Most existing methods tend to employ a greedy strategy, directly heading to the target point with the highest current benefit while neglecting global coverage path optimization. This short-sighted strategy often leads to unnecessary repeated exploration and resource waste, especially in large-scale environments. Although attempts have been made to solve the globally optimal path using the Traveling Salesman Problem (TSP), the complexity of solving the TSP increases dramatically with the number of explored viewpoints, making it difficult to implement in practical applications.

[0005] To address the problems of high computational resource consumption and low exploration efficiency in current methods, this invention aims to provide a lightweight and efficient autonomous exploration method and system for covering three-dimensional scenes. Summary of the Invention

[0006] In order to solve the technical problems existing in the background art, the present invention aims to provide a three-dimensional environment coverage method for online sub-region updating and hierarchical planning.

[0007] To solve the technical problem, the technical solution of the present invention is as follows:

[0008] A method for online sub-region updating and hierarchical planning of 3D environment coverage, the specific technical solution of which is as follows:

[0009] Step S1: Real-time localization and mapping: Using the lidar and inertial measurement unit mounted on the ground unmanned platform, the SLAM (real-time localization and mapping) algorithm is used to scan and map the current environment, while acquiring the pose information of the unmanned platform, and using the UFOMap mapping algorithm to generate a three-dimensional voxel map of the current environment.

[0010] Step S2: Rapid Environment Preprocessing: Rapid preprocessing of the environment area of ​​the current map, mainly including incremental frontier point extraction, sparse topology road map construction and viewpoint generation;

[0011] Step S3: Sub-region division: The entire unknown environment to be covered is initially divided into several sub-region units. Each sub-region unit stores the basic environmental information of that unit, namely the Sub-region Information Structure (SIS), which is used for subsequent exploration planning. Each sub-region unit has a hierarchical structure and is dynamically decomposed according to the volume ratio of the unknown area within the unit. At the same time, the Sub-region Information Structure is also continuously updated to adapt to the dynamic changes of the unknown area.

[0012] Step S4: Global Coverage Path Generation: Based on the divided sub-regions, for sub-regions with viewpoints within them, the entry order of these sub-regions is arranged by solving the asymmetric traveling salesman problem to determine the access priority of each sub-region at time t;

[0013] Step S5: Local exploration path optimization: Guided by the global coverage path, the access order of each viewpoint in a sub-region within a certain range from the unmanned platform is optimized by solving the asymmetric traveling salesman problem again.

[0014] Step S6: Motion planning: Based on the access order of the viewpoints, the next viewpoint to be visited is sent to the motion planning module, prompting the robot to move to the target to explore;

[0015] Step S7: Check if there are any unvisited viewpoints on the map. If no viewpoints are found, end the exploration; otherwise, repeat the above steps.

[0016] Furthermore, using the lidar and inertial measurement unit mounted on the ground unmanned platform, the SLAM algorithm is used to scan and map the current environment, while simultaneously acquiring the pose information of the unmanned platform. The UFOMap mapping algorithm is then used to generate a 3D voxel map of the current environment. Based on the generated 3D voxel map of the current environment, the mapped environmental area is rapidly preprocessed, including incremental front point extraction, sparse topological road map construction, and viewpoint generation, thus obtaining effective information about the current environment.

[0017] Furthermore, the input of the SLAM algorithm is the point cloud data at the current moment and the vehicle acceleration, angular velocity and angular acceleration information measured by the inertial measurement unit, and the output is the robot's pose estimation and point cloud map; the input of the UFOMap algorithm is the pose estimation and the point cloud data at the current moment, and the output is the three-dimensional voxel map of the explored part of the environment.

[0018] Furthermore, the rapid preprocessing of the currently mapped environmental area specifically includes:

[0019] Incremental extraction of frontier points: Obtain the set of all voxels whose occupancy values ​​have changed from UFOMap, check the state of each voxel, and if a voxel is in a free state and its neighboring voxels are in an unknown state, mark the voxel as a frontier voxel. At the same time, during the exploration process, check whether each frontier voxel located in the map update area is still valid and remove invalid frontier voxels.

[0020] Constructing a sparse topology map: Nodes in the topology map represent key points in the environment, and edges represent paths connecting adjacent nodes. During the exploration process, candidate nodes are generated through random sampling. Sampling nodes that meet the conditions will undergo collision checks. Collision checks are performed using ray tracing operations of UFOMap and then added to the topology map. The weight of the edge is set to the distance between two nodes. All nodes are stored in a KD tree structure to improve computational efficiency.

[0021] Viewpoint generation: For each front point, find the node closest to the front in the topology graph and use that node as the viewpoint.

[0022] Furthermore, the initial division of the entire unknown environment to be covered specifically includes:

[0023] Sub-region hierarchical recursive partitioning: In the initial stage of exploration, the entire environment to be explored is divided into multiple sub-region units along the axis-aligned boundary, with each unit initially having a hierarchy of l. i =1, the sub-region will serve as the basic unit for subsequent exploration and planning tasks; each sub-region unit can be recursively divided using a quadtree structure. Initially, the sub-region unit is large, but as the exploration progresses, the volume of the unknown region decreases, and the original sub-region is further subdivided. i The sub-region of level will become l i+1 The hierarchy involves dividing each sub-region unit into smaller sub-regions as the hierarchy increases, thereby capturing the dynamically changing environment during the exploration process.

[0024] Sub-region information structure: Each sub-region unit contains its hierarchical information. i Axis-aligned border range, known area volume percentage R knownVP of all viewpoints in the sub-region, and C of the centroid of all viewpoints. avg The distance L between each sub-region cost Key information used to dynamically update the status of sub-regions;

[0025] Sub-region online update: During exploration, the environment map is continuously updated. Sub-region units intersecting with the map update area are recalculated and their information structure is updated. If the proportion of known area volume in a sub-region exceeds a certain threshold but has not reached the maximum level, the update will be applied. max This will further divide the area into smaller sub-regions, and simultaneously update the path lengths between these sub-regions. max The calculation rules are as follows:

[0026]

[0027] in, and This represents the maximum and minimum allowed side lengths of the subregion. This indicates the rounding up operation.

[0028] Furthermore, by addressing the asymmetric traveling salesman problem, the entry order into each sub-region is arranged, specifically including:

[0029] The global coverage path problem is transformed into an asymmetric traveling salesman problem by constructing a distance cost matrix M between each sub-region. cost Using operations research and optimization solvers such as OR-Tools, a global path with the shortest distance that sequentially enters all sub-regions can be calculated. Only sub-regions with a greater than 0 internal viewpoints are included in the global coverage path calculation process. The distance cost matrix M cost The calculations for each element are as follows:

[0030]

[0031] in This represents the distance cost between sub-regions i and j. This cost has already been calculated in the online sub-region update process in step S3, so it can be filled in directly. N act This represents the total number of sub-regions with a greater than 0 internal viewpoint count;

[0032]

[0033] in, This represents the distance cost from the current position p0 of the unmanned platform to the sub-region k, where p0 is obtained through the pose estimation output by the SLAM algorithm.

[0034] Finally, the standard traveling salesman problem needs to be transformed into an asymmetric traveling salesman problem to optimize the solution speed:

[0035]

[0036] The path length cost in the above process is calculated through the constructed sparse topology graph; specifically, the A* search algorithm is used to search the topology graph to find a collision-free path connecting the search start point and the target point, and the path length cost between the two points can be obtained by calculating the length of the path.

[0037] The constructed cost matrix M cost The input is optimized in OR-Tools. The solver output is the index order in which these sub-regions are entered, denoted as . S i G represents the index of a sub-region, therefore, * This is the shortest path that sequentially covers all sub-regions.

[0038] Furthermore, by addressing the asymmetric traveling salesman problem, the access sequence of various viewpoints within a sub-region of a certain distance from the unmanned platform is optimized, including:

[0039] Based on the entry order G of the obtained sub-regions * From G * Select the distance from the robot radius r max N within near Divide the region into sub-regions and obtain the viewpoints of these sub-regions: Let N be the total number of N. v One viewpoint, among which This represents the j-th viewpoint within the i-th sub-region. The viewpoint is searched and computed in the sparse topological path graph. and viewpoint The distance and cost between And the current position p0 of the unmanned platform reaches the viewpoint Distance and cost :

[0040]

[0041]

[0042] For M cost All elements in the first row are filled with 0 to construct the asymmetric traveling salesman problem:

[0043]

[0044] The cost matrix M cost The solution is optimized by inputting it into OR-Tools, which is affected by the global coverage path G.* Due to limitations, this optimization problem adds constraints, namely the viewpoint. Should be compared to viewpoint The viewpoints are visited first. This constraint can be set in OR-Tools. Therefore, the viewpoint access order obtained after solving will satisfy the above constraint while ensuring the shortest overall access path length.

[0045] Furthermore, the next viewpoint to be visited is sent to the motion planning module to guide the robot to the exploration target, specifically including:

[0046] Based on the obtained viewpoint access order, the viewpoint that should be accessed at the current time t is selected as the target point, and the target point is sent to the path planning module. This module is responsible for planning a path from the robot to the target point and effectively avoiding obstacles during the journey.

[0047] The specific operation includes two parts: global path planning and local path planning. Global path planning uses the A* algorithm, combined with localization information and sparse topology graph search, to obtain the overall path. Local path planning uses point cloud data to obtain obstacle location information and applies the DWA dynamic window algorithm to achieve local obstacle avoidance during robot movement, so as to ensure that the robot can safely and smoothly reach the target area.

[0048] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement a three-dimensional environment overlay method for online sub-region updating and hierarchical planning as described above.

[0049] A computer-readable storage medium storing a computer program that, when executed by a processor, implements a three-dimensional environment coverage method for online sub-region updating and hierarchical planning as described above.

[0050] Compared with the prior art, the advantages of the present invention are as follows:

[0051] This invention proposes a 3D environment coverage method through online sub-region updates and hierarchical planning, addressing the problems of high computational cost and low coverage efficiency in current exploration methods. First, this method proposes a fast environment preprocessing approach that only requires examining already updated boundary regions in the map to extract frontier points, significantly reducing computational overhead. Second, during the exploration process, this method continuously constructs a sparse topological graph of the environment, discretizing the continuous space and storing topological nodes in a KD-tree, facilitating subsequent path search and node lookup, and avoiding expensive path search operations performed on the 3D voxel map.

[0052] Secondly, a hierarchical sub-region structure partitioning method is proposed. This method divides the entire environment to be explored into several sub-region units and allows these units to undergo adaptive decomposition to capture the dynamic changes in unknown areas. During the exploration process, each sub-region unit stores its internal environmental information through a Sub-Region Information Structure (SIS), which is autonomously updated as the exploration progresses to support subsequent exploration planning.

[0053] Furthermore, utilizing a hierarchical planning strategy, the environmental coverage problem is first modeled as a traveling salesman problem, solving for global coverage paths based on sub-region units. This problem considers the total path length required to access these sub-regions. Compared to the greedy strategies of common exploration methods, the proposed method provides a long-term perspective for subsequent exploration planning, thus avoiding short-sightedness in the exploration process. Guided by the global coverage path, local exploration paths are refined, further determining the next target to be explored. This coarse-to-fine decision-making method reduces the difficulty of solving the optimization problem and improves exploration efficiency.

[0054] Finally, the proposed method was validated in both simulation and real-world environments, and the experimental results confirmed its efficiency and feasibility. Attached Figure Description

[0055] Figure 1 The present invention provides a flowchart of a three-dimensional environment coverage method for online sub-region updating and hierarchical planning;

[0056] Figure 2 Visualization of environmental pretreatment results;

[0057] Figure 3 A schematic diagram of the sub-region hierarchical structure;

[0058] Figure 4 , one A schematic diagram of a region division with only two levels;

[0059] Figure 5 A diagram representing hierarchical planning;

[0060] Figure 6 Figure 1 shows the experimental results in a real-world scenario. Detailed Implementation

[0061] The specific implementation of the present invention is described below with reference to embodiments:

[0062] It should be noted that the structures, proportions, sizes, etc. shown in this specification are only used to complement the content disclosed in the specification for those skilled in the art to understand and read, and are not intended to limit the conditions under which the present invention can be implemented. Any modifications to the structure, changes in the proportions, or adjustments to the size, without affecting the effects and objectives that the present invention can produce, should still fall within the scope of the technical content disclosed in the present invention.

[0063] Furthermore, the terms such as "upper," "lower," "left," "right," "middle," and "one" used in this specification are merely for clarity of description and are not intended to limit the scope of the invention. Any changes or adjustments to their relative relationships, without substantially altering the technical content, should also be considered within the scope of the invention.

[0064] Example 1:

[0065] Figure 1 This is a flowchart of a three-dimensional environment coverage method for online sub-region updating and hierarchical planning provided by the present invention.

[0066] Figure 2 This is a visualization of the environmental preprocessing. During exploration, voxel states are incrementally checked to extract leading voxels, while new nodes are continuously sampled around the unmanned platform to expand the topology map. The node on the topology map closest to the leading voxel will be used as the viewpoint.

[0067] Figure 3 This is a schematic diagram of the sub-region hierarchy. Sub-region units can be further subdivided to represent unknown regions using smaller units, up to the maximum level l. max。

[0068] Figure 4 This is a schematic diagram of a region partitioning scheme with only two levels. If the known voxel ratio in a sub-region exceeds a threshold, the initial region will be further subdivided. Regions without a viewpoint within them will be considered invalid sub-regions and will not be included in subsequent global coverage planning.

[0069] Figure 5 This diagram illustrates the hierarchical planning. Arrows represent global coverage paths G* that sequentially enter all valid sub-regions, while dashed lines represent local paths optimized under the guidance of the global path. The local paths will sequentially visit r... max All viewpoints within.

[0070] Figure 6 The results of the experiment are shown in a real-world scenario. The top left corner of the image shows the unmanned experimental platform. Figure 6 The right side shows the actual view of the garage. Figure 6The image below uses point clouds to represent the environmental results of the exploration, curves to represent the exploration trajectory of the unmanned platform, and dots to represent the starting point of the exploration.

[0071] like Figure 1 As shown, this invention proposes a 3D environment coverage method through online sub-region updating and hierarchical planning. This method is verified and tested under a Robot Operating System (ROS). The test platform is as follows: Figure 6 As shown, the specific operation steps of this method are as follows:

[0072] Step S1: Use the point cloud data at the current moment and the vehicle acceleration, angular velocity and angular acceleration information measured by the inertial measurement unit as input information for the SLAM algorithm to obtain the pose estimation information of the unmanned platform, and use the UFOMap algorithm to generate a three-dimensional voxel map of the known environment.

[0073] Step S2: Rapid environment preprocessing extracts data information from the current environment for subsequent exploration and planning. This process includes three parts: incremental extraction of frontier points, construction of a sparse topology roadmap, and viewpoint generation. A schematic diagram of the environment preprocessing is shown below. Figure 2 As shown. The specific processing method is as follows:

[0074] (1) Incremental extraction of frontier points: Obtain the set of all voxels whose occupancy values ​​have changed from UFOMap and check the state of each voxel. If a voxel is in a free state and its neighboring voxels are in an unknown state, then mark the voxel as a frontier voxel. At the same time, during the exploration process, check whether each frontier voxel located in the map update area is still valid and remove invalid frontier voxels. This method only needs to check a small number of voxels in the map, which significantly reduces the workload of frontier point extraction.

[0075] (2) Constructing a sparse topological roadmap: Nodes in the topological graph represent key points in the environment, and edges represent paths connecting adjacent nodes. During the exploration process, candidate nodes are generated through random sampling. Sampling nodes that meet the conditions undergo collision checking, which is performed using ray tracing operations in UFOMap, and then added to the topological graph. The weight of the edge is set to the distance between two nodes. All nodes are stored in a KD-tree structure to improve computational efficiency.

[0076] (3) Viewpoint generation: For each front point, find the node closest to the front in the topology graph and use that node as the viewpoint. The purpose of viewpoint generation is to find a suitable location for environmental perception near the front point so that subsequent exploration and coverage can be achieved.

[0077] Step S3: Sub-region division. The entire exploration environment is divided into several sub-regions. The specific operation is as follows:

[0078] (1) Sub-region hierarchical recursive division: In the initial stage of exploration, the entire environment to be explored is divided into several sub-region units along the axis-aligned boundary, and the initial hierarchy of each unit is l. i = 1, these sub-regions will serve as the basic units for subsequent exploration and planning tasks. Each sub-region unit can be recursively divided using a quadtree structure. Initially, the sub-region units are relatively large, but as exploration progresses, the volume of the unknown region decreases, and the original sub-regions are further subdivided. Originally in l i The sub-region of level will become l i +1 The hierarchy involves dividing each sub-region into smaller sub-regions as the hierarchy increases, thereby capturing the dynamically changing environment during exploration, such as... Figure 3 and Figure 4 As shown.

[0079] (2) Sub-region information structure: Each sub-region unit contains its hierarchical information l i Axis-aligned border range, known area volume percentage R known VP of all viewpoints in the sub-region, and C of the centroid of all viewpoints. avg The distance L between each sub-region cost This information, including key details, is used to dynamically update the status of sub-regions and provide data support for subsequent exploration planning.

[0080] (3) Online Sub-region Update: During the exploration process, the environment map is continuously updated. Sub-region units intersecting with the map update area need to be recalculated to update their information structure. If the proportion of known area volume in a sub-region exceeds a certain threshold but has not reached the maximum level... max This will be further divided into smaller sub-regions. Simultaneously, the path lengths between these sub-regions will be updated to ensure the effectiveness and accuracy of the exploration plan. max The calculation rules are as follows:

[0081]

[0082] in, and This represents the maximum and minimum allowed side lengths of the subregion. This indicates the rounding up operation.

[0083] Step S4: Generate global coverage paths to obtain the optimal order for entering each sub-region. The specific steps are as follows:

[0084] The global coverage path problem is transformed into an asymmetric traveling salesman problem. This is achieved by constructing a distance cost matrix M between each sub-region. costUsing operations research and optimization solvers such as OR-Tools, a global path with the shortest distance that sequentially enters all sub-regions can be calculated, such as... Figure 5 As indicated by the middle arrow segment. Only sub-regions with a greater than 0 internal viewpoints are included in the global coverage path solution process. Distance cost matrix M cost The calculations for each element are as follows:

[0085]

[0086] in This represents the distance cost between sub-region i and sub-region j. This cost has already been calculated in the online sub-region update process in step S3, so it can be filled in directly. N act This represents the total number of sub-regions with a greater than 0 internal viewpoint count.

[0087]

[0088] in, This represents the distance cost from the current position p0 of the unmanned platform to the sub-region k. p0 is obtained through pose estimation output by the SLAM algorithm in step S1.

[0089] Finally, the standard traveling salesman problem needs to be transformed into an asymmetric traveling salesman problem to optimize the solution speed:

[0090]

[0091] The path length cost in the above process is calculated using the sparse topology graph constructed in step S2. Specifically, the A* search algorithm is used to search the topology graph for a collision-free path connecting the search start point and the target point. The path length cost between the two points can be obtained by calculating the length of this path.

[0092] The constructed cost matrix M cost The input is optimized in OR-Tools. The solver output is the index order in which these sub-regions are entered, denoted as . S i This represents the index of a sub-region. Therefore, G * This is the shortest path that sequentially covers all sub-regions.

[0093] Step S5: Local exploration path optimization, which optimizes the access order of various viewpoints in a sub-region within a certain range. The specific operation is as follows:

[0094] According to the entry order G of the sub-regions obtained in step S4 * From G * Select the distance from the robot radius r max N withinnear Divide the region into sub-regions and obtain the viewpoints of these sub-regions:

[0095] Let N be the total number of N. v One viewpoint, among which This represents the j-th viewpoint within the i-th sub-region. Similar to step S5, viewpoints are searched and computed in the sparse topological road map. and viewpoint The distance and cost between And the current position p0 of the unmanned platform reaches the viewpoint Distance and cost :

[0096]

[0097]

[0098] For M cost All elements in the first row are filled with 0 to construct the asymmetric traveling salesman problem:

[0099]

[0100] The cost matrix M cost The solution is then optimized using OR-Tools. Unlike step S5, this is affected by the global coverage path G. * Due to limitations, this optimization problem adds constraints, namely the viewpoint. Should be compared to viewpoint It is visited first. This is because the global coverage path G obtained from the solution is... * In the solution, the i-th sub-region is visited before the (i+1)-th sub-region. This constraint can be set in OR-Tools, so the viewpoint access order obtained after solving will satisfy the above constraint while ensuring the shortest overall access path length, such as... Figure 5 As shown by the dashed line segment.

[0101] Step S6: Motion planning section, prompting the robot to explore the target point, the specific operation is as follows:

[0102] Based on the viewpoint access order obtained in step S5, the viewpoint that should be visited at time t is selected as the target point, and the target point is sent to the path planning module. This module is responsible for planning a path from the robot to the target point and effectively avoiding obstacles during the journey. The specific operations include two parts: global path planning and local path planning. Global path planning uses the A* algorithm, combined with localization information and sparse topology graph search, to obtain the overall path. Local path planning utilizes point cloud data to obtain obstacle location information and applies the DWA (Dynamic Window) algorithm to achieve local obstacle avoidance during robot movement, ensuring that the robot can safely and smoothly reach the target area.

[0103] Step S7: Check if there is a viewpoint in the current environment. If there is no viewpoint, end the exploration. Otherwise, repeat the above steps until there is no viewpoint.

[0104] Figure 6 The results are shown in the experimental setting of an underground parking garage. The garage scene is 50 m × 50 m in size and includes parked vehicles and the narrow areas formed between them. During the experiment, the robot's maximum speed was limited to 1 m / s. Figure 6 The image demonstrates the explored space represented using 3D point cloud representation. The curves in the image represent the robot's trajectory, and the dots indicate the starting points of the trajectories. Throughout the exploration process, there were no redundant paths, and the robot explored 5673.25 m in 175.53 seconds. 3 The vehicle moved 157.53 m. This method completely explored the entire garage, leaving no area unexplored, thus achieving effective exploration of the unknown environment.

[0105] Example 2:

[0106] This embodiment provides a terminal device, which includes a processor and a memory. The memory stores a computer program, which includes program instructions. The processor executes the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or other general-purpose processors, 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, discrete hardware components, etc. It is the computing and control core of the terminal, suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions to achieve a corresponding method flow or corresponding function. The processor described in this embodiment can be used in the operation of a three-dimensional environment coverage method for online sub-region updating and hierarchical planning, including the following steps:

[0107] Step S1: Real-time localization and mapping: Using the lidar and inertial measurement unit mounted on the ground unmanned platform, the SLAM (real-time localization and mapping) algorithm is used to scan and map the current environment, while acquiring the pose information of the unmanned platform, and using the UFOMap mapping algorithm to generate a three-dimensional voxel map of the current environment.

[0108] Step S2: Rapid Environment Preprocessing: Rapid preprocessing of the environment area of ​​the current map, mainly including incremental frontier point extraction, sparse topology road map construction and viewpoint generation;

[0109] Step S3: Sub-region division: The entire unknown environment to be covered is initially divided into several sub-region units. Each sub-region unit stores the basic environmental information of that unit, namely the Sub-region Information Structure (SIS), which is used for subsequent exploration planning. Each sub-region unit has a hierarchical structure and is dynamically decomposed according to the volume ratio of the unknown area within the unit. At the same time, the Sub-region Information Structure is also continuously updated to adapt to the dynamic changes of the unknown area.

[0110] Step S4: Global Coverage Path Generation: Based on the divided sub-regions, for sub-regions with viewpoints within them, the entry order of these sub-regions is arranged by solving the asymmetric traveling salesman problem to determine the access priority of each sub-region at time t;

[0111] Step S5: Local exploration path optimization: Guided by the global coverage path, the access order of each viewpoint in a sub-region within a certain range from the unmanned platform is optimized by solving the asymmetric traveling salesman problem again.

[0112] Step S6: Motion planning: Based on the access order of the viewpoints, the next viewpoint to be visited is sent to the motion planning module, prompting the robot to move to the target to explore;

[0113] Step S7: Check if there are any unvisited viewpoints on the map. If no viewpoints are found, end the exploration; otherwise, repeat the above steps.

[0114] Example 3:

[0115] This embodiment provides a storage medium, specifically a computer-readable storage medium (Memory), which is a memory device in a terminal device used to store programs and data. It is understood that the computer-readable storage medium here can include both the built-in storage medium in the terminal device and extended storage media supported by the terminal device. The computer-readable storage medium provides storage space that stores the terminal's operating system. Furthermore, this storage space also stores one or more instructions suitable for loading and execution by a processor. These instructions can be one or more computer programs (including program code). It should be noted that the computer-readable storage medium here can be high-speed RAM or non-volatile memory, such as at least one disk storage device.

[0116] One or more instructions stored in a computer-readable storage medium can be loaded and executed by a processor to implement the corresponding steps of the three-dimensional environment coverage method for online sub-region updating and hierarchical planning in the above embodiments; one or more instructions in the computer-readable storage medium are loaded by the processor and executed as follows:

[0117] Step S1: Real-time localization and mapping: Using the lidar and inertial measurement unit mounted on the ground unmanned platform, the SLAM (real-time localization and mapping) algorithm is used to scan and map the current environment, while acquiring the pose information of the unmanned platform, and using the UFOMap mapping algorithm to generate a three-dimensional voxel map of the current environment.

[0118] Step S2: Rapid Environment Preprocessing: Rapid preprocessing of the environment area of ​​the current map, mainly including incremental frontier point extraction, sparse topology road map construction and viewpoint generation;

[0119] Step S3: Sub-region division: The entire unknown environment to be covered is initially divided into several sub-region units. Each sub-region unit stores the basic environmental information of that unit, namely the Sub-region Information Structure (SIS), which is used for subsequent exploration planning. Each sub-region unit has a hierarchical structure and is dynamically decomposed according to the volume ratio of the unknown area within the unit. At the same time, the Sub-region Information Structure is also continuously updated to adapt to the dynamic changes of the unknown area.

[0120] Step S4: Global Coverage Path Generation: Based on the divided sub-regions, for sub-regions with viewpoints within them, the entry order of these sub-regions is arranged by solving the asymmetric traveling salesman problem to determine the access priority of each sub-region at time t;

[0121] Step S5: Local exploration path optimization: Guided by the global coverage path, the access order of each viewpoint in a sub-region within a certain range from the unmanned platform is optimized by solving the asymmetric traveling salesman problem again.

[0122] Step S6: Motion planning: Based on the access order of the viewpoints, the next viewpoint to be visited is sent to the motion planning module, prompting the robot to move to the target to explore;

[0123] Step S7: Check if there are any unvisited viewpoints on the map. If no viewpoints are found, end the exploration; otherwise, repeat the above steps.

[0124] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0125] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0126] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0127] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0128] The preferred embodiments of the present invention have been described in detail above. However, the present invention is not limited to the above embodiments. Within the scope of knowledge possessed by those skilled in the art, various changes can be made without departing from the spirit of the present invention.

[0129] Many other changes and modifications can be made without departing from the concept and scope of this invention. It should be understood that this invention is not limited to the specific embodiments, and the scope of this invention is defined by the appended claims.

Claims

1. A three-dimensional environment coverage method for online sub-region updating and hierarchical planning, characterized in that, The method includes: Based on the obtained valid information about the current environment, the entire unknown environment to be covered is initially divided into multiple sub-regional units; Based on the divided sub-region units, for sub-regions with viewpoints within them, the order of entry into each sub-region is arranged by solving the asymmetric traveling salesman problem; Guided by the global coverage path, the access order of various viewpoints in a sub-region within a certain range from the unmanned platform is optimized by solving the asymmetric traveling salesman problem. Based on the access order of each viewpoint, the next viewpoint to be accessed is sent to the motion planning module to guide the robot to the exploration target; Check if there are any unvisited viewpoints on the map. If no viewpoints are found, end the exploration; otherwise, repeat the above steps. Using the lidar and inertial measurement unit mounted on the ground unmanned platform, the SLAM algorithm is used to scan and map the current environment, while acquiring the pose information of the unmanned platform. The UFOMap mapping algorithm is then used to generate a 3D voxel map of the current environment. Based on the generated 3D voxel map of the current environment, the mapped environmental area is rapidly preprocessed, including incremental front point extraction, sparse topology road map construction, and viewpoint generation, thus obtaining effective information about the current environment. The SLAM algorithm takes the current point cloud data and the vehicle acceleration, angular velocity, and angular acceleration information measured by the inertial measurement unit as inputs, and outputs the robot's pose estimation and point cloud map; the UFOMap algorithm takes the pose estimation and the current point cloud data as inputs, and outputs a 3D voxel map of the explored part of the environment. The rapid preprocessing of the currently mapped environmental area specifically includes: Incremental extraction of frontier points: Obtain the set of all voxels whose occupancy values ​​have changed from UFOMap, check the state of each voxel, and if a voxel is in a free state and its neighboring voxels are in an unknown state, mark the voxel as a frontier voxel. At the same time, during the exploration process, check whether each frontier voxel located in the map update area is still valid and remove invalid frontier voxels. Constructing a sparse topology map: Nodes in the topology map represent key points in the environment, and edges represent paths connecting adjacent nodes. During the exploration process, candidate nodes are generated through random sampling. Sampling nodes that meet the conditions will undergo collision checks. Collision checks are performed using ray tracing operations of UFOMap and then added to the topology map. The weight of the edge is set to the distance between two nodes. All nodes are stored in a KD tree structure to improve computational efficiency. Viewpoint generation: For each front point, find the node closest to that front point in the topology graph and use that node as the viewpoint.

2. The three-dimensional environment coverage method for online sub-region updating and hierarchical planning according to claim 1, characterized in that, The initial division of the entire unknown environment to be covered specifically includes: Sub-region hierarchical recursive partitioning: In the initial stage of exploration, the entire environment to be explored is divided into multiple sub-region units along the axis-aligned boundary, with each unit initially having a hierarchy of l. i =1, the sub-region will serve as the basic unit for subsequent exploration and planning tasks; each sub-region unit can be recursively divided using a quadtree structure. Initially, the sub-region unit is large, but as the exploration progresses, the volume of the unknown region decreases, and the original sub-region is further subdivided. i The sub-region of level will become l i+1 The hierarchy involves dividing each sub-region unit into smaller sub-regions as the hierarchy increases, thereby capturing the dynamically changing environment during the exploration process. Sub-region information structure: Each sub-region unit contains its hierarchical information. i Axis-aligned border range, known area volume percentage R known VP of all viewpoints in the sub-region, and C of the centroid of all viewpoints. avg The distance L between each sub-region cost Key information used to dynamically update the status of sub-regions; Sub-region online update: During exploration, the environment map is continuously updated. Sub-region units intersecting with the map update area are recalculated and their information structure is updated. If the proportion of known area volume in a sub-region exceeds a certain threshold but has not reached the maximum level, the update will be applied. max This will further divide the area into smaller sub-regions, and simultaneously update the path lengths between these sub-regions. max The calculation rules are as follows: ； in, and This represents the maximum and minimum allowed side lengths of the subregion. This indicates the rounding up operation.

3. The three-dimensional environment coverage method for online sub-region updating and hierarchical planning according to claim 1, characterized in that, By addressing the asymmetric traveling salesman problem, the order of entry into each sub-region is arranged, specifically including: The global coverage path problem is transformed into an asymmetric traveling salesman problem by constructing a distance cost matrix M between each sub-region. cost Using the OR-Tools operations research solver, a global path with the shortest distance that sequentially enters all sub-regions can be calculated. Only sub-regions with a greater than 0 internal viewpoints are included in the global coverage path calculation process. The distance cost matrix M... cost The calculations for each element are as follows: ； in, N represents the cost of the distance between subregions i and j. act This represents the total number of sub-regions with a greater than 0 internal viewpoint count; ； in, This represents the distance cost from the current position p0 of the unmanned platform to the sub-region k, where p0 is obtained through pose estimation output by the SLAM algorithm. Finally, the standard traveling salesman problem needs to be transformed into an asymmetric traveling salesman problem to optimize the solution speed: ； The path length cost in the above process is calculated through the constructed sparse topology graph. Specifically, the A* search algorithm is used to search the topology graph to find a collision-free path connecting the search start point and the target point. The path length cost between the two points can be obtained by calculating the length of the path. The constructed cost matrix M cost The input is optimized in OR-Tools. The solver output is the index order in which these sub-regions are entered, denoted as . S i G represents the index of a sub-region, therefore, * This is the shortest path that sequentially covers all sub-regions.

4. The three-dimensional environment coverage method for online sub-region updating and hierarchical planning according to claim 1, characterized in that, By addressing the asymmetric traveling salesman problem, the access order of various viewpoints within a sub-region of a certain distance from the unmanned platform is optimized, including: Based on the entry order G of the obtained sub-regions * From G * Select the distance r from the robot radius max N within near Divide the region into sub-regions and obtain the viewpoints of these sub-regions: Let N be the total number of N. v One viewpoint, among which This represents the j-th viewpoint within the i-th sub-region. The viewpoint is searched and computed in the sparse topological path graph. and viewpoint The distance and cost between And the current position p0 of the unmanned platform reaches the viewpoint Distance and cost : ； ； For M cost All elements in the first row are filled with 0 to construct the asymmetric traveling salesman problem: ； The cost matrix M cost The solution is optimized by inputting it into OR-Tools, which is affected by the global coverage path G. * The constraints imposed on the optimization problem, namely the viewpoint, are added. Should be compared to viewpoint The viewpoints are visited first. This constraint can be set in OR-Tools. Therefore, the viewpoint access order obtained after solving will satisfy the above constraint while ensuring the shortest overall access path length.

5. The three-dimensional environment coverage method for online sub-region updating and hierarchical planning according to claim 1, characterized in that, The next viewpoint to be visited is sent to the motion planning module to guide the robot toward the exploration target, specifically including: Based on the obtained viewpoint access order, the viewpoint that should be accessed at the current time t is selected as the target point, and the target point is sent to the path planning module. This module is responsible for planning a path from the robot to the target point and effectively avoiding obstacles during the journey. The specific operation includes two parts: global path planning and local path planning. Global path planning uses the A* algorithm, combined with localization information and sparse topology graph search, to obtain the overall path. Local path planning uses point cloud data to obtain obstacle location information and applies the DWA dynamic window algorithm to achieve local obstacle avoidance during robot movement, so as to ensure that the robot can safely and smoothly reach the target area.

6. A computer device, characterized in that, The system includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements a three-dimensional environment overlay method for online sub-region updating and hierarchical planning as described in any one of claims 1 to 5.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements a three-dimensional environment coverage method for online sub-region updating and hierarchical planning as described in any one of claims 1 to 5.

Images