A method and system for constructing and optimizing environmental maps based on ultra-wideband (UWB)

Abstract

The present invention provides a method and system for constructing and optimizing environmental maps based on ultra-wideband (UWB), which is suitable for extreme environments. First, the system initializes a two-dimensional grid map and collects signal data through UWB devices. When determining the LOS path, the corresponding grid is defined as a "free" state, otherwise it is marked as "occupied". For NLOS features, the obstacle position is mapped to the map. Next, the map is refined by a mean filter, a morphological filter and a region growing algorithm to improve the accuracy of obstacle representation. Finally, the optimized map information is used to adjust the positioning algorithm weights through the weighted least squares method (WLS) to improve positioning accuracy. The advantages of the present invention include: it can effectively construct environmental obstacle information, which is suitable for navigation after refinement, and can still construct maps normally in dark light and smoke environments. The system relies only on UWB sensors to achieve precise positioning in obstacle environments.

Description

Technical Field

[0001] The present invention relates to a simultaneous localization and mapping (SLAM) technology, and in particular to a method and system for constructing and optimizing an environment map based on ultra-wideband (UWB). Background Art

[0002] Simultaneous localization and mapping (SLAM) is a complex technology that plays a vital role in autonomous mobile robots and unmanned navigation. It requires the system to be able to simultaneously build a map in an unknown environment and accurately determine its real-time position. Currently, LiDAR systems and camera vision systems are common types of external sensors used for SLAM. However, the frequency range used by LiDAR and camera light sources is close to the visible light frequency range in the electromagnetic spectrum. Therefore, in challenging scenes such as those filled with smoke or containing a large number of mirrors, the signal-to-noise ratio of these optical measurement devices is relatively low, making SLAM algorithms that rely on such devices prone to failure, significantly reducing the functional performance of autonomous systems.

[0003] Currently, navigation tasks in high-temperature environments are a crucial challenge. For example, during fire rescue operations, it is necessary to quickly locate trapped people through smoke and plan the safest evacuation route. In underground mines, high temperatures, dust, and complex geological structures place higher demands on technology. In the kitchen environment of the catering industry, especially those in large commercial kitchens, the continuous high temperature and heavy oil smoke have a significant impact on autonomous systems. In recent years, given the stability of radio frequency (RF) signals in harsh and extreme environmental conditions, a variety of studies have explored the use of RF bands to replace traditional dedicated sensors for SLAM. Some new RF devices, such as ultra-wideband radar (UWB radar) and micro-Doppler radar, have initially been able to complete SLAM tasks in challenging extreme environments, but their relatively high cost has hindered their widespread application in industrial production and experiments.

[0004] Due to the frequency characteristics of radio frequency signals and the relative cost-effectiveness of anchor-tag systems, robot positioning based on UWB systems has been widely studied. However, completing the mapping algorithm based on UWB systems remains a challenge.

[0005] Mapping based on UWB radar falls under the category of feature point matching and mapping. UWB radar transmits signals at specific intervals and collects echo signals. These echo signals are processed to calculate the time difference between the transmitted and received signals, and then infer distance information based on the signal propagation speed. By utilizing the time of arrival (TOA) triangulation principle, two sets of these distance measurements can be used on a two-dimensional plane to preliminarily determine the coordinates of the nearest obstacle. The unsupervised DBSCAN clustering algorithm then effectively clusters obstacles into a single cluster, identifying obstacle entities. As the obstacle moves, the positions of the obstacles are continuously optimized, completing the algorithm and constructing a map. This method demonstrates the feasibility of UWB signal mapping in complex and harsh environments. However, unlike traditional UWB anchor-tag systems, UWB radar cannot be integrated into communication devices such as mobile phones and watches, hindering its widespread application. UWB radar is also expensive and costly. UWB radar positioning still relies on point cloud feature point recognition, resulting in low positioning accuracy.

[0006] The system integrates sonar and lidar to build maps in visually obstructed environments. The system is equipped with a smoke sensor. When the smoke concentration in the environment exceeds a preset threshold, causing the effectiveness of conventional optical sensing methods to decrease, the system automatically switches to relying on sonar data for spatial perception and information acquisition. Under this strategy, sonar detection data is converted through a corresponding algorithm to replace the point cloud information lost by the lidar due to smoke obstruction. This conversion process ensures that even when the lidar's direct observation capability is obstructed, effective perception and precise positioning of the environment can be maintained, allowing for continuous map construction. However, sonar contains less information and still relies on lidar data. At the same power, the sonar transmission distance is much shorter than that of UWB signals, resulting in a smaller mapping range.

[0007] In extremely harsh indoor environments, such as those covered by smoke, dim light, and glass, traditional optical measurement equipment like lidar and cameras have relatively low signal-to-noise ratios. This makes SLAM algorithms that rely on these devices prone to failure, significantly reducing the functional performance of autonomous systems and leading to mission failure. New sensor devices like UWB radar and sonar still rely on feature point matching algorithm frameworks, which results in inadequate measurement information, limited measurement range, low accuracy, and high system costs.

[0008] Therefore, map construction in extremely harsh indoor environments requires: 1. Fully utilizing signal information to complete map construction; 2. Unaffected by harsh environments and possessing good robustness; 3. The equipment must be integrable and capable of use in large indoor scenarios.

[0009] It should be noted that the information disclosed in the above background technology section is only used to understand the background of this application, and therefore may include information that does not constitute prior art known to ordinary technicians in this field. Summary of the Invention

[0010] The main purpose of the present invention is to solve the problems existing in the above background technology and provide a method and system for constructing and optimizing environmental maps based on ultra-wideband (UWB).

[0011] To achieve the above object, the present invention adopts the following technical solutions:

[0012] In a first aspect of the present invention, a method for constructing and optimizing an environment map based on ultra-wideband (UWB) comprises the following steps:

[0013] S1. Map Construction: Initialize a two-dimensional grid map and collect signal data from the environment using a UWB device. When a clear line-of-sight (LOS) path exists between the base station and the tag, all grid cells along the path are defined as "free." If the communication result indicates a non-LOS state, the relevant grid area is preliminarily classified as "occupied," assuming that an obstacle may exist in the "occupied" area. If a non-line-of-sight (NLOS) feature is detected in a known "free" grid cell, the impact of the obstacle is mapped onto the grid map, forming a visual representation of the obstacle's location.

[0014] S2. Map Refinement: The initially constructed map is optimized using binary filtering. This includes: applying a mean filter to remove noise and smooth obstacle edges to reduce misjudgments of areas initially marked as "occupied." A morphological filter is used to further verify and adjust the initially marked obstacle shapes through dilation and erosion operations to ensure the accuracy of obstacle boundaries. A region growing algorithm is applied to refine obstacle boundaries, merging adjacent similar regions. The initially marked "occupied" status is verified and adjusted to achieve accurate representation of obstacles in the map.

[0015] S3. Map and positioning algorithm optimization: Utilizing the optimized map information, the weighted least squares method (WLS) is used to adjust the positioning algorithm weights to mitigate the adverse effects of NLOS data. Different weights are assigned to each ranging data point based on its NLOS identification result, with lower weights given to data significantly affected by NLOS and higher weights given to LOS data, thereby improving positioning accuracy. This results in an optimized map and positioning results.

[0016] In a second aspect of the present invention, a system for constructing and optimizing an environment map based on ultra-wideband (UWB) includes:

[0017] The UWB positioning system includes multiple UWB base stations and at least one UWB tag carried on the mobile robot, which is used to measure the distance between the base station and the tag to provide raw data for positioning;

[0018] An IMU sensor, mounted on the mobile robot, is used to collect motion data of the mobile robot and assist in UWB positioning;

[0019] A mobile robot, used to carry UWB tags and IMU sensors to move around the environment and take measurements;

[0020] The processing module communicates with the UWB base station and the UWB tag, and is used to execute the method for constructing and optimizing the environment map based on ultra-wideband UWB.

[0021] In a third aspect of the present invention, a computer-readable storage medium stores a computer program, wherein when the computer program is executed by a processor, the method for constructing and optimizing an environment map based on ultra-wideband (UWB) is implemented.

[0022] In a fourth aspect of the present invention, a computer program product includes a computer program, which implements the method for constructing and optimizing an environment map based on ultra-wideband (UWB) when executed by a processor.

[0023] The present invention has the following beneficial effects:

[0024] This paper proposes a method and system for constructing and optimizing environmental maps based on ultra-wideband (UWB). This system utilizes ultra-wideband (UWB) technology to effectively construct and optimize environmental maps in extremely harsh indoor environments, such as those affected by smoke, dim lighting, or glass reflections. This method accurately constructs environmental obstacle information through communication between a base station and tags, and refines this information to enable subsequent navigation tasks. Even in adverse conditions such as dim lighting and smoke, the system maintains the continuity and accuracy of map construction.

[0025] This invention improves the accuracy and reliability of map construction by mapping UWB-based NLOS information onto a two-dimensional grid map. A binary filtering algorithm is used to correct the two-dimensional grid map constructed by UWB, further improving map quality. Furthermore, the real-time prior grid map further modifies the positioning algorithm, optimizing positioning accuracy.

[0026] Using only UWB as an external sensor, the system achieves precise positioning even in obstructed environments, demonstrating excellent robustness. Furthermore, the present invention implemented a real-world experiment using a Mecanum car model, validating the algorithm's applicability and effectiveness in diverse environments. Overall, the present invention provides a cost-effective, highly accurate, adaptable, and easily integrated UWB map construction and optimization solution.

[0027] Other beneficial effects of the embodiments of the present invention will be further described below. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 This is a framework diagram of the mapping algorithm according to an embodiment of the present invention.

[0029] Figure 2 This is a visual representation of the NLOS mapping algorithm according to an embodiment of the present invention.

[0030] Figure 3 Schematic diagram of the UWB mapping process according to an embodiment of the present invention.

[0031] Figure 4 This is a diagram of the physical process of UWB mapping in an embodiment of the present invention.

[0032] Figure 5 This is a diagram of the actual result of UWB mapping in an embodiment of the present invention.

[0033] Figure 6 This is a diagram showing the multi-robot mapping effect of an embodiment of the present invention. DETAILED DESCRIPTION

[0034] The following is a detailed description of the embodiments of the present invention. It should be emphasized that the following description is only exemplary and is not intended to limit the scope of the present invention and its application.

[0035] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of such features. In the description of the embodiments of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0036] The present invention provides a method for constructing and optimizing an environmental map based on ultra-wideband (UWB). The method primarily comprises: 1. Mapping the results of UWB-based NLOS information determination onto a two-dimensional grid map to complete map construction; 2. Correcting the UWB-constructed two-dimensional grid map using a binary filtering algorithm; and 3. Further correcting the positioning algorithm using a real-time prior grid map. The present invention achieves real-time mapping of unknown environments in extremely harsh environments. A physical experiment using a Mecanum car model was implemented, enabling algorithm validation in diverse environments.

[0037] See Figure 1 The embodiment of the present invention provides a method for constructing and optimizing an environment map based on ultra-wideband (UWB), comprising the following steps:

[0038] Step S1. Map construction: Initialize a two-dimensional grid map and collect signal data in the environment through the UWB device; when there is a clear line-of-sight (LOS) path between the base station and the tag, all grid cells on the path are defined as "free"; if the communication result shows a non-LOS state, the relevant grid area is preliminarily classified as "occupied", assuming that the "occupied" area may have obstacles; if NLOS characteristics are found in a known "free" grid cell, the impact of the obstacle is mapped to the grid map to form a visual representation of the obstacle location.

[0039] In some embodiments, step S1 specifically includes:

[0040] Initialize the map: Create a blank two-dimensional grid map to record environmental information;

[0041] Data collection: Use UWB devices to collect signal data between base stations and tags in the environment;

[0042] LOS path identification: Detects whether there is a LOS path between the base station and the tag to determine whether the signal can propagate in a straight line;

[0043] Definition of free space: For each grid cell on the LOS path, mark it as "free", indicating that there are no obstacles in the area;

[0044] Non-LOS processing: For grid areas in non-LOS state, a conservative strategy is adopted to mark them as "occupied", indicating that there may be obstacles in the area;

[0045] NLOS feature mapping: For NLOS features detected in “free” grid cells, analyze and infer the actual location of the obstacle and visualize the location on the map;

[0046] Map state update: Based on the results of LOS path identification, non-LOS processing, and NLOS feature mapping, the grid state on the map is updated to reflect the current estimated position of obstacles in the environment.

[0047] Step S2. Map refinement: The initially constructed map is optimized using binary filtering, including: applying a mean filter to remove noise and smooth obstacle edges to reduce misjudgments of areas initially marked as "occupied"; using a morphological filter to further verify and adjust the initially marked obstacle shapes through dilation and erosion operations to ensure the accuracy of obstacle boundaries; applying a region growing algorithm to refine obstacle boundaries, merge adjacent similar regions, and verify and adjust the initially marked "occupied" status to achieve accurate representation of obstacles in the map;

[0048] In some embodiments, in step S2, applying a mean filter to remove noise and smooth the edges of obstacles specifically includes: selecting an N×N window to calculate the neighborhood average of each pixel in the map; calculating the average value of all pixels within the N×N window; replacing the original value of the center pixel of the window with the calculated average value; repeating the above process for each pixel on the map until all pixels of the entire map are processed.

[0049] In some embodiments, in step S2, further verifying and adjusting the initially marked obstacle shape using a morphological filter through dilation and erosion operations specifically includes:

[0050] First, a dilation operation is applied around the boundary of the initially identified obstacle to expand the marked obstacle area; then an erosion operation is applied to the map after the dilation operation to shrink the obstacle area and eliminate isolated noise points; the dilation operation is followed by the erosion operation to fill small bright crack-like discontinuities.

[0051] In some embodiments, in step S2, the region growing algorithm is applied to refine the obstacle boundary, merge adjacent similar regions, verify and adjust the "occupied" status of the preliminary mark, and achieve accurate representation of obstacles in the map, specifically including: selecting pixels marked as obstacles in the map as seed pixels; performing similarity evaluation on adjacent pixels around the seed pixels to determine whether they meet a predetermined similarity standard with the seed pixels; merging adjacent pixels that meet the similarity standard into the obstacle region to which the seed pixel belongs; performing iterative expansion, repeating the evaluation and merging operations until no more adjacent pixels that meet the conditions can be added to the obstacle region and the obstacle region reaches a saturated state.

[0052] Step S3. Map and positioning algorithm optimization: Using the optimized map information, the positioning algorithm weights are adjusted through the weighted least squares (WLS) method to reduce the adverse effects of non-line-of-sight (NLOS) data. Different weights are assigned to each ranging data point based on its NLOS identification result, with data that is more affected by NLOS being given a lower weight and LOS data being given a higher weight to improve positioning accuracy. This results in an optimized map and positioning result.

[0053] In some embodiments, step S3 specifically includes: assigning different weights to each data point based on the NLOS identification result of each ranging data point; wherein, a lower weight is assigned to the data point that is more affected by NLOS, and a higher weight is assigned to the LOS data point; constructing a weighted least squares (WLS) problem to optimize the positioning algorithm, and determining the position by solving the problem; using optimization techniques such as the Newton-Gauss method or the Levenberg-Marquardt method to minimize the objective function; feeding back the occupancy information in the grid map into the positioning algorithm to enhance the accuracy and reliability of positioning; adjusting the weight based on the prior knowledge of obstacles and non-line-of-sight conditions in the grid map to adapt to the correction requirements of non-line-of-sight errors; using map information to achieve more accurate error suppression and performance optimization, thereby improving positioning accuracy.

[0054] In a preferred embodiment, for the constructed weighted least squares (WLS) problem, the weight of each ranging data point in the positioning process is dynamically adjusted according to the ratio of the total number of grids between the base station and the tag to the number of obstacle grids; an influencing factor is introduced to dynamically adjust the weight of the positioning algorithm according to the distribution of obstacles and non-line-of-sight conditions in the map to adapt to different environmental conditions.

[0055] An embodiment of the present invention also provides a system for constructing and optimizing an environmental map based on ultra-wideband (UWB), comprising: a UWB positioning system, comprising multiple UWB base stations and at least one UWB tag mounted on a mobile robot, for measuring the distance between the base station and the tag, and providing raw data for positioning; an IMU sensor mounted on the mobile robot, for collecting motion data of the mobile robot, and assisting UWB positioning; a mobile robot, for carrying the UWB tag and the IMU sensor, for moving in the environment and performing measurements; a processing module, for communicating with the UWB base station and the UWB tag, for executing the method for constructing and optimizing an environmental map based on ultra-wideband (UWB).

[0056] The present invention provides a system that can accurately construct and optimize environmental maps even in extremely harsh indoor environments, such as those exposed to smoke, dim light, or glass. Utilizing ultra-wideband (UWB) technology, the system effectively constructs environmental obstacle information through communication between a base station and tags, and through refinement processing, makes the map information suitable for subsequent navigation tasks. Even in adverse conditions such as dim light and smoke, the system maintains the continuity and accuracy of map construction. This method improves the accuracy and reliability of map construction by mapping UWB non-line-of-sight (NLOS) information onto a two-dimensional grid map. The binary filtering algorithm employed further optimizes the map and enhances the accuracy of obstacle representation. Furthermore, the real-time constructed prior grid map further modifies the positioning algorithm and optimizes positioning accuracy. The system of the present invention uses only UWB as an external sensor and can perform precise positioning even in obstructed environments, demonstrating excellent robustness. Furthermore, the system has been subjected to physical experiments on a Mecanum car model, verifying the applicability and effectiveness of the algorithm in various environments. In general, the present invention provides a cost-effective, highly accurate, adaptable and easy-to-integrate UWB map construction and optimization solution, which can achieve efficient and reliable map construction and positioning in a variety of complex environments.

[0057] The following further describes specific embodiments of the present invention and examples of algorithm implementation thereof.

[0058] System composition

[0059] Software system composition

[0060] The software system runs on the Ubuntu 18.04 (a Linux operating system mainly for desktop applications) operating system and the ROS-melodic (open source robot operating system) software environment.

[0061] ROS Melodic, with its powerful modular architecture, rich software package collection, and flexible programming interface, greatly simplifies the development, testing, and deployment of robotics software. It provides a series of plug-and-play tools and libraries, including sensor data processing, motion control, path planning, and visual recognition, enabling developers to quickly build and iterate complex robotics functions. Through seamless integration with Ubuntu 18.04, ROS Melodic can fully utilize the resource management, network support, and graphical interface provided by the operating system, creating a highly integrated and efficient development environment for robotics researchers and engineers.

[0062] Hardware system composition

[0063] The experiment uses a four-wheeled mobile robot RIKEBOT. RIKEBOT is a wheeled robot equipped with four omnidirectional Mecanum wheels, a LeTMC-520RGBD depth camera, a RP-LiDAR 2D laser radar, a NoopLoop:LinkTrack UWB device, an MPU6050 IMU sensor, and computing hardware. Jetson Nano TM , with a quad-core ARM processor and 16GB of memory. A bracket was installed on the robot to ensure that there were no obstacles blocking the UWB signal. A tag was mounted on the robot's bracket at the same height as the four base stations in the environment.

[0064] UWB

[0065] Nooploop's LinkTrack UWB is an advanced positioning system based on ultra-wideband (UWB) technology. It offers installation support on multiple operating systems, including Ubuntu, allowing users to quickly build and configure the system based on their needs. In addition to the hardware, Nooploop also provides supporting software tools and algorithms to enable advanced positioning services such as track tracking, environmental mapping, and data analysis.

[0066] Visualization Platform

[0067] The mapping algorithm uses RVIZ as its visualization platform. RVIZ is a powerful 3D visualization tool within the Robot Operating System (ROS), primarily used for displaying and debugging various data and status information in robotic systems. Built on the Qt framework, it's an open platform designed to provide users with an intuitive way to observe and understand the dynamics of robots and their environments. RVIZ is an integral part of the ROS ecosystem, greatly facilitating the visualization of complex information during robot development, testing, and maintenance, and serves as a powerful auxiliary tool for robotics developers and researchers.

[0068] Range-SLAM is based on the UWB mapping algorithm framework and runs on the Ubuntu 18.04 operating system and ROS-melodic software environment. It is compiled and run on the main control machine of the hardware control system. Figure 2 It consists of a value filtering module and a positioning algorithm correction module. Its functions include constructing the surrounding environment, refining obstacle information, and completing environmental positioning under obstacles.

[0069] Mapping Algorithm

[0070] A method for building an environmental map using UWB is proposed. It mainly consists of three parts: the first part realizes the correspondence between non-line-of-sight propagation and obstacles in the two-dimensional map; the second part uses an efficient binary filtering algorithm to further refine obstacles in the map; the third part further optimizes positioning through the constructed map to achieve a closed loop of simultaneous positioning and mapping. The process framework of the mapping algorithm is as follows Figure 1 shown.

[0071] 1. NLOS Mapping

[0072] In the UWB mapping system, when there is a clear line of sight (LOS) path between the base station and the tag, it means that there are no obstacles blocking the spatial transmission channel between the two, and the signal can propagate in a straight line without interference. Based on this ideal situation, all grid cells projected by the light along this path can be defined as "free", denoted as {M line} LOS =C free ,These grids usually represent open and unobstructed areas of space.

[0073] Similarly, if the communication results between the base station and the tag show that they are not in a LOS state, this means that there is some form of physical obstacle on the expected path, hindering the direct propagation of the signal. In this case, for such non-line-of-sight (NLOS) scenarios, the relevant grid areas need to be processed more carefully.

[0074] First, when facing grid areas that have not been fully detected or have unclear information, if they are initially judged to be in NLOS state, in order to ensure the robustness and security of the system, a conservative strategy is generally adopted to temporarily classify these grids as "occupied" state, that is, {M line} NLOS =C occupy , assuming that there may be potential obstacles in these areas that affect signal propagation.

[0075] Secondly, if a portion of a known "free" grid cell is found to actually exhibit NLOS characteristics, this does not mean that the grid itself contains physical obstacles, but rather that there are obstacles on the path extending outward to the base station or tag. In this case, the core concept of the mapping framework is reflected - by converting complex NLOS identification results into an intuitive geospatial representation, that is, mapping the impact of obstacles onto the grid map, forming a visual representation of the obstacle location.

[0076] The specific expression can be expressed in mathematical formula as formula (4.1)

[0077]

[0078] In summary, this framework cleverly uses NLOS information to indirectly infer the location of actual obstacles and displays these insights by occupying or releasing grid cells, transforming the originally abstract radio wave propagation environment into an easy-to-understand and easy-to-operate grid map model, greatly enhancing the understanding and optimization of environmental perception and positioning performance.

[0079] Figure 2 A visual representation of the NLOS mapping algorithm is shown.

[0080] 2. Binary Filtering Optimization Map

[0081] In preliminary applications of the NLOS mapping algorithm, a raw, unoptimized map was generated. However, this map has several inherent issues: First, the initially constructed map often contains noise in the data collected by the sensor, which can lead to inaccurate occupancy status identification in certain areas; for example, there may be temporary misidentifications or missed detections. Second, due to the limited resolution of the sensor, the map may display blurred transition boundaries, especially at the edges of objects. For example, the edges of objects may be partially incorrectly marked as navigable space.

[0082] To further enhance and optimize the original image, an adaptive binary filtering technique was implemented. A binary filter is an image processing technique used to perform a binary filtering operation on an image. The primary function of this filter is to extract specific structural information about obstacles in the image while simultaneously eliminating noise. In this paper, the binary filter consists of three components: a mean filter, a morphological filter, and a region growing algorithm.

[0083] Mean filter

[0084] The implementation of the mean filter is to calculate the average value of all pixels in the N×N window, and then replace the original value of the central pixel with the average value calculated by the window, see formula (4.2)

[0085]

[0086] This approach is effective for removing low-frequency noise and constitutes a straightforward linear filtering technique.

[0087] Morphological filter

[0088] The dilation operation is used to expand the obstacle area, and the erosion operation is used to shrink the obstacle area and eliminate isolated noise points. After using the averaging filter to remove small black point noise, the closing operation is applied, that is, dilation followed by erosion, to fill small bright crack-like discontinuities. Specifically, Equations (4.3) and (4.4) represent the erosion and dilation operations respectively:

[0089]

[0090] Region Growing Algorithm

[0091] In the implementation of the region growing algorithm, adjacent pixels are checked in detail to see if they meet the predefined similarity criteria. Once these pixels meet the criteria, they are merged into the region currently being constructed. Subsequently, the iterative process of evaluation and expansion continues, applying the same review and growth steps to the newly included pixels until the entire region reaches saturation, which means that no more qualified new pixels can be found, at which point the algorithm naturally ends. Generally speaking, region growing algorithms are mostly applied to grayscale images. Here, adjustments are made so that they can be applied to filtering in binary images, thereby better obtaining the required obstacle area. The specific formula is shown in Equation (4.5):

[0092]

[0093] Here, R represents the region occupied by the obstacle in the current map, and N represents the set of neighboring nodes surrounding R. The function d represents the degree of similarity between region R and the grid point. T represents a threshold. If the similarity measure is less than the threshold, the grid point becomes occupied and becomes part of the obstacle region R. Here, the improved region growing algorithm applied to binary images demonstrates remarkable effectiveness in optimizing obstacle region segmentation.

[0094] 3. Optimization of UWB Positioning Using Prior Maps

[0095] A positioning method based on weighted least squares (WLS) is introduced to mitigate the adverse effects of non-line-of-sight (NLOS) data. Its characteristic is that it assigns different weights to each ranging data point based on its NLOS identification result. This results in a lower weighting of data significantly affected by NLOS when solving the positioning problem, while a higher weighting is given to data with line-of-sight (LOS) distance. This processing method is expected to make the positioning algorithm more adaptable and robust in complex and changing non-line-of-sight (NLOS) environments, thereby significantly improving positioning accuracy.

[0096] The following WLS question will be constructed:

[0097]

[0098] where β i is the weight of the ranging measurement, which is related to the score of the NLOS judgment measurement, and λ is a constant coefficient. Minimization is achieved by optimization techniques such as the Newton-Gauss method or the Levenberg-Marquardt method.

[0099] The occupancy grid map not only serves as the basis for environmental representation, but also its stored occupancy information will be actively fed back to the positioning link to enhance the accuracy and reliability of positioning. When combined with the map information obtained by UWB, the weight β in the original positioning algorithm is i The setting of will be changed to accommodate the need to correct for non-line-of-sight errors. Formula (3.8) will be revised to Formula (3.9) to better reflect and utilize the prior knowledge of obstacles and non-line-of-sight conditions in the environment map, thereby achieving more accurate error suppression and performance optimization in the actual positioning process.

[0100]

[0101] where N Total,i and N Occupy,i They represent the total number of grids and the number of obstacle grids passed on the connecting line between base station i and the tag, respectively, and ζ∈[0,1] indicates the degree to which positioning is affected by the map.

[0102] Algorithm 1: Range-SLAM pseudo code is as follows:

[0103]

[0104] See Figures 3 to 6 , Figure 3 The UWB mapping process of an embodiment of the present invention is shown. Figure 4 The actual process of UWB mapping in an embodiment of the present invention is shown. Figure 5 The actual result of UWB mapping in an embodiment of the present invention is shown. Figure 6 The multi-robot mapping effect of an embodiment of the present invention is demonstrated.

[0105] Compared with the existing technology, the main advantages of UWB map construction and optimization through the present invention are:

[0106] 1. Through the communication between the base station and the tag, the environmental obstacle information can be normally constructed;

[0107] 2. The obstacle information on the map has been refined and can be used for subsequent navigation;

[0108] 3. In extreme dark light and fog conditions, it does not affect the construction of the map.

[0109] 4. The system only uses UWB as an external sensor and can still complete the positioning function well in an obstacle environment.

[0110] An embodiment of the present invention further provides a storage medium for storing a computer program, which at least performs the above method when executed.

[0111] An embodiment of the present invention further provides a control device, comprising a processor and a storage medium for storing a computer program; wherein the processor is configured to execute at least the method described above when executing the computer program.

[0112] An embodiment of the present invention further provides a processor, which executes a computer program and at least performs the method described above.

[0113] The storage medium can be implemented by any type of non-volatile storage device, or a combination thereof. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a magnetic random access memory (FRAM), a flash memory, a magnetic surface memory, an optical disc, or a compact disc read-only memory (CD-ROM); the magnetic surface memory can be a magnetic disk memory or a magnetic tape memory. The storage medium described in the embodiments of the present invention is intended to include, but is not limited to, these and any other suitable types of memory.

[0114] In the several embodiments provided by the present invention, it should be understood that the disclosed systems and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0115] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0116] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0117] Those skilled in the art will appreciate that all or part of the steps of the above-mentioned method embodiments may be implemented by hardware associated with program instructions, and the aforementioned program may be stored in a computer-readable storage medium. When the program is executed, the program executes the steps of the above-mentioned method embodiments. The aforementioned storage medium includes various media that can store program codes, such as mobile storage devices, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical disks.

[0118] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

[0119] The methods disclosed in the several method embodiments provided by the present invention can be arbitrarily combined without conflict to obtain new method embodiments.

[0120] The features disclosed in several product embodiments provided by the present invention can be arbitrarily combined without conflict to obtain new product embodiments.

[0121] The features disclosed in several method or device embodiments provided by the present invention can be arbitrarily combined without conflict to obtain new method embodiments or device embodiments.

[0122] The above is a further detailed description of the present invention in conjunction with specific preferred embodiments, and the specific implementation of the present invention should not be considered to be limited to these descriptions. Those skilled in the art will recognize that, without departing from the scope of the present invention, several equivalent substitutions or obvious variations can be made, and the performance or use of the same should be considered to fall within the scope of protection of the present invention.

Claims

1. A method for constructing and optimizing an environment map based on ultra-wideband (UWB), characterized in that: The following steps are involved: S1. Map Construction: Initialize a 2D grid map and collect signal data from the environment using a UWB device. When a clear line-of-sight (LOS) path exists between the base station and the tag, all grid cells along the path are defined as "free." If the communication result indicates a non-LOS state, the relevant grid area is preliminarily classified as "occupied," assuming that an obstacle may exist in the "occupied" area. If a non-line-of-sight (NLOS) signature is detected in a known "free" grid cell, the impact of the obstacle is mapped onto the grid map, creating a visual representation of the obstacle's location. S2. Map Refinement: Binary filtering is performed on the initially constructed map. This includes applying a mean filter to remove noise and smooth obstacle edges to reduce misidentification of areas initially marked as "occupied." Morphological filters are used to further verify and adjust the shapes of initially marked obstacles through dilation and erosion operations to ensure the accuracy of obstacle boundaries. A region growing algorithm is applied to refine obstacle boundaries, merging adjacent similar regions, and verifying and adjusting the initially marked "occupied" status to achieve accurate representation of obstacles in the map. S3. Map and positioning algorithm optimization: Utilizing the optimized map information, the positioning algorithm weights are adjusted using the weighted least squares method (WLS) to mitigate the adverse effects of NLOS data. Different weights are assigned to each ranging data point based on its NLOS identification result, with data significantly affected by NLOS being given a lower weight and LOS data being given a higher weight, thereby improving positioning accuracy. This results in an optimized map and positioning results. In step S2, the region growing algorithm is applied to refine the obstacle boundaries, merge adjacent similar regions, verify and adjust the initially marked "occupied" status, and achieve accurate representation of obstacles in the map, specifically including: Select pixels marked as obstacles in the map as seed pixels; Performing similarity evaluation on adjacent pixels around the seed pixel to determine whether they meet a predetermined similarity standard with the seed pixel; Merge adjacent pixels that meet the similarity criteria into the obstacle region to which the seed pixel belongs; Iterative expansion, repeated evaluation and merging operations are performed until no more eligible adjacent pixels can be added to the obstacle area and the obstacle area reaches saturation.

2. The method for constructing and optimizing an environment map based on ultra-wideband (UWB) according to claim 1, characterized in that: Step S1 specifically includes: Create a blank 2D raster map ready for recording environmental information; Use UWB devices to collect signal data between base stations and tags in the environment; Detect whether there is a LOS path between the base station and the tag to determine whether the signal can propagate in a straight line; For the grid cells on the LOS path, they are marked as "free", indicating that there are no obstacles in the area; For the grid area in non-LOS state, a conservative strategy is adopted to mark it as "occupied", indicating that there may be obstacles in the area; For NLOS features detected in "free" grid cells, analyze and infer the actual location of the obstacle and visualize the location on the map; Based on the results of LOS path identification, non-LOS processing, and NLOS feature mapping, the grid state on the map is updated to reflect the current estimated positions of obstacles in the environment.

3. The method for constructing and optimizing an environment map based on ultra-wideband (UWB) according to claim 1, wherein: In step S2, applying a mean filter to remove noise and smooth the obstacle edges specifically includes: Select an N×N window to calculate the neighborhood average of each pixel in the map; Calculate the average value of all pixels in the N×N window; Replace the original value of the center pixel of the window with the calculated average value; Repeat the above process for each pixel on the map until all pixels in the map have been processed.

4. The method for constructing and optimizing an environment map based on ultra-wideband (UWB) according to claim 1, wherein: In step S2, the morphological filter is used to further verify and adjust the shape of the initially marked obstacle through dilation and erosion operations, specifically including: First, a dilation operation is applied around the boundary of the initially identified obstacle to expand the marked obstacle area; then an erosion operation is applied to the map after the dilation operation to shrink the obstacle area and eliminate isolated noise points; the dilation operation is followed by the erosion operation to fill small bright crack-like discontinuities.

5. The method for constructing and optimizing an environment map based on ultra-wideband (UWB) according to claim 1, wherein: Step S3 specifically includes: According to the NLOS identification result of each ranging data point, different weights are assigned to each data point; among them, data points that are more affected by NLOS are given lower weights, while LOS data points are given higher weights; Construct a weighted least squares (WLS) problem to optimize the positioning algorithm, and determine the position by solving the problem; Use optimization techniques to minimize the objective function; Feeding occupancy information from the grid map into the positioning algorithm to enhance positioning accuracy and reliability; Adjust the weights based on the prior knowledge of obstacles and non-line-of-sight conditions in the grid map to meet the correction requirements of non-line-of-sight errors; Utilize map information to achieve more precise error suppression and performance optimization, improving positioning accuracy.

6. The method for constructing and optimizing an environment map based on ultra-wideband (UWB) according to claim 5, characterized in that: The optimization techniques include the Newton-Gauss method or the Levenberg-Marquardt method.

7. The method for constructing and optimizing an environment map based on ultra-wideband (UWB) according to claim 5, characterized in that: For the constructed weighted least squares (WLS) problem, the weight of each ranging data point in the positioning process is dynamically adjusted according to the ratio of the total number of grids between the base station and the tag to the number of obstacle grids; an influencing factor is introduced to dynamically adjust the weight of the positioning algorithm according to the distribution of obstacles and non-line-of-sight conditions in the map to adapt to different environmental conditions.

8. A system for constructing and optimizing environmental maps based on ultra-wideband (UWB), characterized in that: include: The UWB positioning system includes multiple UWB base stations and at least one UWB tag carried on the mobile robot, which is used to measure the distance between the base station and the tag to provide raw data for positioning; An IMU sensor, mounted on the mobile robot, is used to collect motion data of the mobile robot and assist in UWB positioning; A mobile robot, used to carry UWB tags and IMU sensors to move around the environment and take measurements; The processing module communicates with the UWB base station and the UWB tag, and is used to execute the method for constructing and optimizing an environment map based on ultra-wideband (UWB) according to claim 1.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method for constructing and optimizing an environment map based on ultra-wideband (UWB) according to any one of claims 1 to 7 is implemented.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the method for constructing and optimizing an environment map based on ultra-wideband (UWB) according to any one of claims 1 to 7 is implemented.

Images