Grid map construction method and system in complex time-varying environment
By adaptively adjusting the position and scale of the grid, the problem of wasted computational resources and optimization conflicts in traditional grid maps in complex and time-varying environments is solved, achieving efficient grid map construction that is suitable for the field of path planning.
Patent Information
- Application Number
- CN202512009575.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-29
- Publication Date
- 2026-02-03
AI Technical Summary
Traditional raster maps struggle to balance computational efficiency and high spatial resolution in complex, time-varying environments, leading to wasted computational resources and conflicts between local optimization and global path planning.
By obtaining the correlation between each grid and its neighboring grids, a grid of interest is defined, and the grid scale factor is adjusted based on a preset amount of interest. Hooke's law is used to obtain the adjustment force, and the grid position is iteratively solved to achieve adaptive adjustment of the grid position to achieve force balance.
It achieves optimal dynamic allocation of computing resources in complex time-varying environments, forms a multi-level raster map, reduces redundant data, and improves data query and analysis efficiency.
Smart Images
Figure CN121453035A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of target navigation technology, specifically relating to a method and system for constructing grid maps in complex time-varying environments. Background Technology
[0002] Raster maps are widely used in various fields, including robot navigation, autonomous driving, drone path planning, and dynamic traffic management. Their main advantage lies in their ability to accurately track, predict, and update dynamic obstacles in real time through rasterized environment modeling and the integration of multi-sensor data, thus supporting path planning and decision-making in dynamic scenarios.
[0003] However, traditional raster map construction uses uniform, fixed-size grids across the entire environment (whether open or complex), such as... Figure 2 As shown, to ensure the accuracy of critical areas (such as narrow passages), a high resolution must be used globally, resulting in a huge number of grid cells and wasted computing resources. Therefore, dynamic grids have emerged. However, in complex scenes, dynamic grids need to find a balance between high spatial resolution and system computational efficiency, balancing fine granularity and computational efficiency. Specifically, grid cells that are too small will lead to high computational overhead, while resolutions that are too large may lose important details. In addition, dynamic grids are based on local real-time updates, which makes it difficult to perfectly integrate with global modeling algorithms (such as global path planning). In large-scale task scenarios (such as long-distance drone flights), global path optimization may conflict with local obstacle avoidance. Summary of the Invention
[0004] This invention addresses the shortcomings of existing technologies by providing a method and system for constructing raster maps in complex time-varying environments, which can balance computational efficiency and high spatial resolution, and avoid conflicts between global optima and local optima.
[0005] This invention provides the following technical solution: Firstly, a method for constructing raster maps in complex time-varying environments is provided, including: The geometric information of the working environment is obtained by collecting data from various sensors, and the working environment is divided into grids. Obtain the correlation between each grid and its neighboring grids, and define the grids with a correlation exceeding a set threshold as the grids of interest for the current grid; The scale factor of each grid is defined based on the preset interest value, and the distance between the current grid and the grid of interest is adjusted accordingly. Based on the distance adjustment scale, the adjustment force of each grid of interest on the current grid is obtained according to Hooke's law, and the position of each grid when all grids reach force equilibrium is solved by iteration, thus completing the construction of the grid map.
[0006] Optionally, the correlation of each grid with its surrounding neighboring grids is obtained by obtaining the correlation of the grid and the grid , and the specific formula is: ; ; ; wherein, is a set initial distance scale factor, and are the values in the grid and the grid , respectively, is a set of values in the surrounding neighboring grids of the grid , is the geodesic distance of and ; denotes the cosine function, denotes the inverse hyperbolic function, denotes the norm, denotes the L2 norm, and the superscript denotes the transpose operation.
[0007] Optionally, the scale factor of the grid is obtained by the following formula: ; wherein, is the scale factor of the grid , is the gradient modulus of the physical quantity of interest in the grid , usually the gradient modulus of stress or temperature, is a set constant to prevent the weight from being zero in the area where the gradient is zero, is a set constant to control the sensitivity of the gradient to the scale factor.
[0008] Optionally, the distance adjustment scale is calculated by the following formula: ; wherein, is the distance adjustment scale of the grid and the grid of interest , is the scale factor of the grid of interest .
[0009] Optionally, the adjustment force of each grid of interest on the current grid is obtained according to Hooke's law based on the distance adjustment scale, and the specific formula is: ; in, For grid Received interest grid Adjustment force For grid and grid of interest Distance adjustment scale, and They are grids and grid of interest The current position vector.
[0010] Optionally, the condition for determining that all grids have reached force balance is: when the movement of all grid nodes is less than a set value or the magnitude of the resultant force on all grids is less than a set threshold, then a force balance state is reached.
[0011] Optionally, the iterative method for determining the position of each grid cell when all grid cells reach force equilibrium specifically involves: in each iteration, all grid cells are updated with adaptive displacement; when all nodes reach force equilibrium, the updated grid cell positions are output, where the adaptive displacement of the grid cell is: ; in, For grid exist The adaptive displacement in the next iteration The relaxation factor is set. For grid exist The resultant force experienced during the next iteration.
[0012] Secondly, a raster map construction system for complex time-varying environments is provided, including: The grid division module is used to acquire geometric information of the working environment through data collected by each sensor and divide the working environment into grids. The region of interest determination module obtains the correlation between each grid and its neighboring grids, and defines the grids with a correlation exceeding a set threshold as the grids of interest for the current grid. The scale adjustment module is used to define the scale factor of each grid based on the preset interest quantity, and thereby obtain the distance adjustment scale between the current grid and the grid of interest. The map generation module is used to adjust the scale based on distance, obtain the adjustment force of each grid of interest on the current grid according to Hooke's Law, and solve the position of each grid when all grids reach force equilibrium through iteration, thus completing the construction of the grid map.
[0013] In a third aspect, a computer device is provided, comprising a processor and a memory; wherein the processor implements the steps of the grid map construction method in a complex time-varying environment according to any one of the first aspect when executing a computer program stored in the memory.
[0014] In a fourth aspect, a computer readable storage medium is provided for storing a computer program; the computer program is executed by a processor to implement the steps of the grid map construction method in a complex time-varying environment according to any one of the first aspect.
[0015] Compared with the prior art, the present application has the following beneficial effects: The present application initially divides the global environment into a plurality of network grids, and then adjusts the positions of the grids adaptively to increase the grid density in a local area where the physical quantity changes dramatically or needs higher precision, and to reduce the grid density in an area where the change is smooth, so as to realize optimal dynamic allocation of computing resources; secondly, the present application naturally forms a multi-level representation of high precision in the local area (near the planned path and near the obstacle) and low precision in the global environment on the same grid map, which can perfectly reconcile the inherent contradiction between high resolution and computing efficiency; in addition, the present application can make the grids conform to the geometric characteristics of the environment by adaptively adjusting the positions of the grids, so that the constructed grid map can be expressed more reasonably; finally, the present application can more effectively organize and store data, reduce redundancy, and make data query and analysis more efficient by dynamically adjusting the granularity of the grids. BRIEF DESCRIPTION OF DRAWINGS
[0016] Figure 1 is a flowchart of the grid map construction method in a complex time-varying environment of the present application; Figure 2 is a grid map after division by fixed-size grids in the prior art; Figure 3 is a grid comparison map after division by fixed-size grids and adaptive adjustment of the positions of the grids of the present application. DETAILED DESCRIPTION
[0017] The present application will be further described below with reference to the accompanying drawings. The following examples are only used to more clearly illustrate the technical solutions of the present application, and cannot be used to limit the protection scope of the present application. It should be noted that the terms "comprise" and any variations thereof in the specification and claims of the present application and the above-described drawings are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device comprising a series of steps or units does not have to be limited to only those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.
[0018] Example 1 As Figure 1 shown, a grid map construction method in a complex time-varying environment is provided, comprising the following steps: Step S1: Obtain the geometric information of the working environment through the data acquisition of each sensor, and divide the working environment into grids.
[0019] Any complex working environment is globally planned, divided into small cells (usually three-dimensional space grids) according to the prior art, set to 0.1 meter resolution, and each cell not only represents static objects, but also records properties such as speed and acceleration of dynamic targets.
[0020] The divided cells include initial global modeling and coarse grid generation, and the geometric information of the working environment is obtained through sensors such as laser radar and depth camera. Specifically, the following sub-steps are included: S11: environmental perception and geometric modeling, all entities such as robots, obstacles, boundaries, etc. are constructed into geometric models (such as CAD models or surface grids reconstructed by point clouds). S12: Generate an initial background grid to discretize the entire computational domain (including all geometric entities) with individual grids. This initial grid is global and static, and is used to establish the baseline of the entire physical field. The data in the grid is multi-dimensional.
[0021] Step S2: Obtain the correlation of each grid and its surrounding adjacent grids, and define the grids with correlation exceeding the set threshold as the grids of interest of the current grid.
[0022] In this embodiment, the threshold of correlation can be set according to expert experience. The surrounding adjacent grids are obtained by K-neighbor algorithm. The set of values in the surrounding adjacent grids is represented as: ; Wherein, is the number of surrounding adjacent grids.
[0023] The specific way to obtain the correlation is: ; ; ; Wherein, is the correlation of grid and grid , is the set initial distance scale factor, usually set to 1, and are the values in grid and grid , such as speed or position, is the grid The set formed by the values in the surrounding grid cells. for and The geodesic distance; Represents the cosine function. Represents the inverse hyperbolic function. Represents the norm, Denotes the L2 norm, superscript This indicates the transpose operation.
[0024] Step S3: Define the scale factor of each grid based on the preset interest quantity, and use it to obtain the distance adjustment scale between the current grid and the grid of interest.
[0025] The scale factor of the raster is obtained by the following formula: ; in, For grid The scale factor, For grid The gradient magnitude of the physical quantity of interest, typically the gradient magnitude of stress or temperature, corresponds to a larger scale factor as the gradient increases. To prevent the weights from being zero in regions where the gradient is zero, a constant is set to ensure the overall smoothness of the raster map. This is a constant set to control the gradient's sensitivity to the scale factor. The scale factor is the engine driving the entire refinement process, determining the pace of raster adjustments.
[0026] While keeping the mesh topology (number of cells and nodes, connection relationships) unchanged, the positions of the grid nodes are moved to make them cluster in regions of drastic physical field changes (high gradient regions) and sparse in regions of gradual change. The refinement level can be gradually decreased as the distance from the robot increases.
[0027] The distance adjustment scale is calculated using the following formula: ; in, For grid and grid of interest Distance adjustment scale, For grids of interest The scale factor, the distance adjustment scale, is inversely proportional to the average of the size factors of the two grids.
[0028] Step S4: Based on the distance adjustment scale, obtain the adjustment force of each grid of interest on the current grid according to Hooke's Law, and solve the position of each grid when all grids reach force equilibrium through iteration, thus completing the construction of the grid map.
[0029] Step S4 is a dynamic adjustment of the grid by a distance scale factor, so as to adapt the grid position. After the grid division in step S1, a grid network is obtained, and each grid is subjected to an adjustment force from the adjacent grids. According to Hooke's law, the adjustment force of a grid of interest on a grid is:
[0030] The resultant force on a grid is the resultant force of all the grids of interest, and is specifically:
[0031] In this embodiment, the determination condition for all the grids to reach a force balance is that when the movement amount of all the grid nodes is less than a set value or the size of the resultant force on all the grids is less than a set threshold, the force balance state is reached. That is, the iteration stopping condition is to reach the force balance state, and of course, when the maximum iteration round is reached, the iteration is also stopped.
[0032] The size of the resultant force on all the grids is less than a set threshold, and the ideal state of this set threshold is 0, but in the actual calculation process, the set threshold of the resultant force can only be close to 0, and can be determined according to expert experience. The positions of all the grids when reaching the force balance are solved by iteration, which is a nonlinear equation group about the grid positions (because depends on the position). An iterative method is usually used for solving.
[0033] In each iteration, all the grids are updated in position by an adaptive displacement amount, and the updated grid position is output when all the nodes reach the force balance.
[0034] The adaptive displacement amount of the grid is: is the adaptive displacement amount of the grid in the th iteration, is a set relaxation factor, which is used to ensure the stability of the iteration and prevent the nodes from moving too fast to cause entanglement between the grid networks, is the grid In The resultant force experienced by the grid at the next iteration.
[0035] As Figure 3 shown, Figure 3 (a) and (b) in the figure respectively represent the schematic diagram after fixed size grid division and the schematic diagram of using the adaptive dynamic adjustment grid position of the application to divide the grid, through the update before and after the grid, it can be known that the application can change the resolution of the grid map in the way of changing the grid position, so as to be applicable to the field of path planning, such as path planning of unmanned aerial vehicle or robot.
[0036] The application effectively organizes and stores data, reduces redundancy, and makes data query and analysis more efficient by dynamically adjusting the granularity of the grid, increases the grid density in the local area where the physical quantity changes sharply or needs higher precision, and reduces the grid density in the area where the change is gentle to optimize the calculation efficiency and precision.
[0037] Embodiment 2 A grid map construction system in a complex time-varying environment, comprising: A grid division module for obtaining geometric information of a working environment through collected data of each sensor and dividing the working environment into grids; A region of interest determination module for obtaining the correlation of each grid and its surrounding adjacent grids, and defining the grid whose correlation exceeds a set threshold as the grid of interest of the current grid; A scale adjustment module for defining the scale factor of each grid based on a preset quantity of interest, and obtaining the distance adjustment scale of the current grid and the grid of interest based on the scale factor; A map generation module for obtaining the adjustment force of each grid of interest on the current grid based on the distance adjustment scale according to Hooke's law, and solving the position of each grid when all grids reach force balance through iteration to complete the construction of the grid map.
[0038] The more specific process of the above modules can refer to the corresponding content disclosed in the foregoing embodiments, which will not be repeated here.
[0039] Embodiment 3 The application provides a computer device, comprising a processor and a memory; wherein the processor implements the steps of the above-mentioned grid map construction method in a complex time-varying environment when executing the computer program saved in the memory.
[0040] The more specific process of the above method can refer to the corresponding content disclosed in the foregoing embodiments, which will not be repeated here.
[0041] Embodiment 4 The application provides a computer readable storage medium for storing a computer program; the computer program is executed by a processor to implement steps of the grid map construction method in a complex time-varying environment.
[0042] More specific processes of the above method can refer to the corresponding content disclosed in the foregoing embodiments, and will not be described here.
[0043] The embodiments in the specification are described in a progressive manner, and each embodiment focuses on the difference from other embodiments. The same or similar parts of each embodiment can be referred to each other. For the system, device and storage medium disclosed by the embodiments, since it corresponds to the method disclosed by the embodiments, the description is relatively simple, and the relevant part can be referred to the method part.
[0044] Those skilled in the art can clearly understand that the technology in the embodiments of the application can be realized by means of software and necessary general hardware platforms. Based on such understanding, the technical solutions in the embodiments of the application can be embodied in the form of a software product, which can be stored in a storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, etc., and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the method described in the embodiments of the application or some parts of the embodiments.
[0045] The above is only the preferred embodiment of the application, and the protection scope of the application is not limited to the above-mentioned embodiments. Any technical solution falling within the idea of the application shall fall within the protection scope of the application. It should be noted that, for ordinary skilled in the art, some improvements and refinements without departing from the principles of the application shall be considered as the protection scope of the application.
Claims
1. A method for constructing raster maps under complex time-varying environments, characterized in that, include: The geometric information of the working environment is obtained by collecting data from various sensors, and the working environment is divided into grids. Obtain the correlation between each grid and its neighboring grids, and define the grids with a correlation exceeding a set threshold as the grids of interest for the current grid; The scale factor of each grid is defined based on the preset interest value, and the distance between the current grid and the grid of interest is adjusted accordingly. Based on the distance adjustment scale, the adjustment force of each grid of interest on the current grid is obtained according to Hooke's law, and the position of each grid when all grids reach force equilibrium is solved by iteration, thus completing the construction of the grid map.
2. The method for constructing a raster map in a complex time-varying environment according to claim 1, characterized in that, The process of obtaining the correlation between each grid cell and its neighboring grid cells includes obtaining the grid cells. and grid Relevance The specific formula is as follows: ； ； ； in, The initial distance scale factor is set. and They are grids and grid The values inside, For grid The set formed by the values in the surrounding grid cells. for and The geodesic distance; Represents the cosine function. Represents the inverse hyperbolic function. Represents the norm, Denotes the L2 norm, superscript This indicates the transpose operation.
3. The method for constructing a raster map in a complex time-varying environment according to claim 1, characterized in that, The scale factor of the raster is obtained by the following formula: ； in, For grid The scale factor, For grid The gradient mode of the physical quantity of interest, usually the gradient mode of stress or temperature. To prevent the weights from being zero in regions where the gradient is zero, a constant is set. This is a constant used to control the gradient's sensitivity to the scaling factor.
4. The method for constructing a raster map in a complex time-varying environment according to claim 3, characterized in that, The distance adjustment scale is calculated using the following formula: ； in, For grid and grid of interest Distance adjustment scale, For grids of interest The scale factor.
5. The method for constructing a raster map in a complex time-varying environment according to claim 1, characterized in that, The adjustment force of each grid of interest on the current grid is obtained based on the distance adjustment scale and Hooke's law, specifically as follows: ； in, For grid Received interest grid Adjustment force For grid and grid of interest Distance adjustment scale, and They are grids and grid of interest The current position vector.
6. The method for constructing a raster map in a complex time-varying environment according to claim 1, characterized in that, The condition for determining that all grids have reached a force balance is: when the movement of all grid nodes is less than a set value or the magnitude of the resultant force on all grids is less than a set threshold, then a force balance state is reached.
7. The method for constructing a raster map in a complex time-varying environment according to claim 1, characterized in that, The method of iteratively determining the position of each grid cell when all grid cells reach force equilibrium is as follows: In each iteration, all grid cells are updated with adaptive displacement. When all nodes reach force equilibrium, the updated grid cell positions are output. The adaptive displacement of the grid cells is: ； in, For grid exist The adaptive displacement in the next iteration The relaxation factor is set. For grid exist The resultant force experienced during the next iteration.
8. A raster map construction system for complex time-varying environments, characterized in that, include: The grid division module is used to acquire geometric information of the working environment through data collected by each sensor and divide the working environment into grids. The region of interest determination module obtains the correlation between each grid and its neighboring grids, and defines the grids with a correlation exceeding a set threshold as the grids of interest for the current grid. The scale adjustment module is used to define the scale factor of each grid based on the preset interest quantity, and thereby obtain the distance adjustment scale between the current grid and the grid of interest. The map generation module is used to adjust the scale based on distance, obtain the adjustment force of each grid of interest on the current grid according to Hooke's Law, and solve the position of each grid when all grids reach force equilibrium through iteration, thus completing the construction of the grid map.
9. A computer device, characterized in that, It includes a processor and a memory; wherein, when the processor executes the computer program stored in the memory, it implements the steps of the grid map construction method under complex time-varying environments as described in any one of claims 1-7.
10. A computer-readable storage medium, characterized in that, Used to store computer programs; when the computer programs are executed by a processor, they implement the steps of the grid map construction method under complex time-varying environments as described in any one of claims 1-7.