Robot path dynamic optimization generation method

Abstract

The invention relates to a robot path dynamic optimization generation method, which comprises the following steps: acquiring real-time environment observation information in a local visual field of a robot; updating the global reference path at a fixed frequency based on the global reference path in combination with real-time environment observation information and robot motion constraints; and generating a local reference path by adopting a DWA local path algorithm according to the real-time environment observation information and the updated global reference path. Compared with the prior art, by dynamically updating the local reference path and optimally generating the local path, the real-time performance and accuracy of path planning are improved, the local path generation effect is improved, the robot can avoid obstacles earlier and more accurately, meanwhile, resource consumption is reduced, and remarkable superiority is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of robot path planning, and in particular to a method for dynamically optimizing and generating a robot path. Background Art

[0002] Wheeled mobile robots have been widely used in logistics, services and other fields due to their flexibility and stability. Path planning and obstacle avoidance are the basis for robots to achieve movement and many other tasks, and are one of the indispensable key performance indicators representing robot intelligence. Path planning refers to determining a feasible path from the starting point to the target location for the robot, while obstacle avoidance refers to how to make the robot avoid obstacles during the path planning process to ensure safe and smooth arrival at the target location.

[0003] Existing technical solutions usually use A* or other path generation algorithms to generate a global path, and then use DWA or other local path algorithms to generate local paths so that the robot can adjust and avoid obstacles according to the real-time environment. This approach still has some problems in practical applications. First, since the global path is generated based on a static environment, when the environment changes, the global path may no longer be applicable and needs to be replanned, which will cause certain delays and resource consumption. Secondly, although the existing local path generation algorithms, such as DWA (dynamic window approach), can be adjusted according to the real-time environment, the generated path may still not be ideal in complex environments or narrow spaces, and obstacles cannot be completely avoided, affecting the safety and efficiency of the robot. In addition, when updating the local global path, the existing technical solutions cannot obtain the latest environmental information in a timely and accurate manner, resulting in inaccurate local reference paths, which affects the navigation and obstacle avoidance effects of the robot. Summary of the invention

[0004] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and to provide a robot path dynamic optimization generation method, which can solve the problems of static global path, insufficient local path generation and inaccurate local and global path updates in the prior art, and improve the robot's navigation and obstacle avoidance capabilities in complex environments.

[0005] The object of the present invention can be achieved by the following technical solution: A method for dynamically optimizing and generating a robot path comprises the following steps:

[0006] S1. Obtain real-time environmental observation information within the robot's local field of view;

[0007] S2, based on the global reference path, combined with real-time environmental observation information and robot motion constraints, the global reference path is updated at a fixed frequency;

[0008] S3. Based on the real-time environmental observation information and the updated global reference path, the DWA local path algorithm is used to generate a local reference path.

[0009] Furthermore, the step S1 specifically obtains real-time environmental observation information through sensors carried by the robot.

[0010] Furthermore, the sensor includes a laser radar, a depth camera, an ultrasonic radar, and an obstacle avoidance radar.

[0011] Furthermore, the real-time environmental observation information is specifically environmental information within a set distance range in front of the robot.

[0012] Furthermore, the environmental information includes information about the location, shape, and distance of obstacles.

[0013] Furthermore, the global reference path in step S2 is specifically based on a pre-constructed map, based on obstacle information on the map, and through a path search algorithm, to obtain the path coordinates from the current position to the target position on the grid map.

[0014] Furthermore, the robot motion constraints in step S2 include maximum speed and minimum turning radius.

[0015] Furthermore, the fixed-frequency updating of the global reference path in step S2 specifically involves regenerating the global reference path using an A* algorithm at a preset time interval.

[0016] Furthermore, the step S3 specifically selects a path point on the global reference path within the current visible field of view as the target point of the current travel, and uses the DWA local path algorithm to generate a local reference path from the current position to the target point.

[0017] Furthermore, the target point of the current walking is adjusted according to the real-time environmental observation information. If there is no obstacle within the set distance range in front of the robot, a path point closest to the current position is directly selected from the global reference path;

[0018] If there is an obstacle within a set distance in front of the robot, a path point is selected from the global reference path whose distance to the obstacle meets a preset threshold.

[0019] Compared with the prior art, the present invention has the following advantages:

[0020] 1. The present invention updates the reference path of the global reference path in the local field of view at a fixed frequency according to the observed real-time environment in the local field of view, so as to generate a better local reference path suitable for the current environment for use by the subsequent DWA algorithm. This dynamic update method can obtain the latest environmental information in a timely and accurate manner, making the generated local reference path more accurate, and improving the real-time and accuracy of path planning.

[0021] 2. In the process of local fixed-frequency path search, the present invention selects appropriate target points to generate a path that meets the robot's motion constraints and stays away from nearby obstacles. Especially in narrow road areas, by updating the local global path, the local global reference path is made the center line of the narrow road or the center line of the corner. The local reference path generated in this way can better avoid obstacles, improve the local path generation effect, and improve the safety and efficiency of the robot.

[0022] 3. The present invention uses the DWA local path algorithm to generate a local reference path based on real-time environmental observation information and the updated global reference path, which can avoid obstacles earlier. This ability to avoid obstacles in advance can provide the robot with more obstacle avoidance time and path selection in complex environments or narrow spaces, further improving the safety and efficiency of the robot.

[0023] 4. Since the present invention only updates the global reference path within the local field of view rather than the entire global path, it can reduce the consumption of computing resources, reduce latency, and improve system efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a schematic diagram of the method flow of the present invention. DETAILED DESCRIPTION

[0025] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0026] Example

[0027] like Figure 1 As shown, a method for dynamically optimizing and generating a robot path comprises the following steps:

[0028] S1. Obtain real-time environmental observation information within the robot's local field of view;

[0029] S2, based on the global reference path, combined with real-time environmental observation information and robot motion constraints, the global reference path is updated at a fixed frequency;

[0030] S3. Based on the real-time environmental observation information and the updated global reference path, the DWA local path algorithm is used to generate a local reference path.

[0031] Applying the above scheme to practice, the main contents are:

[0032] 1. Dynamic update of global reference path: In the local field of view of the robot, the global reference path is updated at a fixed frequency according to the observed real-time environment, so as to adjust the global path in time according to the changes in the current environment and avoid static path problems caused by environmental changes. Among them, the global reference path is the robot based on the pre-built map, according to the obstacle information on the map, through the path search algorithm (such as Dijkstra, A*, RRT, RRT*), to determine multiple path coordinates from the current point to the target point on the grid map.

[0033] Real-time environmental information is the current surrounding environment information obtained through various sensors carried by the robot (lidar, depth camera, ultrasonic radar, obstacle avoidance radar, etc.).

[0034] 2. Optimization and generation of local-global paths: In the process of local-global fixed-frequency path search, select appropriate target points, combine obstacle distances, robot motion constraints, etc., and the algorithm for regenerating the path can use A* or other algorithms to generate a more ideal local path and improve the safety and efficiency of the robot.

[0035] It should be noted that after the global reference path is generated, the robot will navigate according to the generated global path points. However, the original global path points are generated based on the pre-built map and do not take into account the robot's current environmental observation data. Therefore, when the robot is walking, the design selects a path point on the global reference path within the current visible field of view as the target point of the current walking. The target point of the local global reference path will be dynamically updated as the robot moves and the field of view changes.

[0036] All reference path points between the robot's current position and the target point of this local global reference path are called the local global path. The local global path is used for the actual local path and trajectory control (such as the reference path of DWA).

[0037] The target point of the local global reference path is adjusted according to the current observation environment within the field of view. If the surrounding area is relatively open, the corresponding point on the global reference path can be used directly; if it is close to an obstacle, fine-tune and select a path point farther away from the obstacle.

[0038] 3. Accurate update of local-global paths: Within the local field of view, the global reference path is updated at a fixed frequency according to the observed real-time environment in order to generate a better local reference path suitable for the current environment, ensuring that the updated local-global path accurately reflects the current environmental information and improves the robot's navigation and obstacle avoidance effects.

[0039] 4. Advance obstacle avoidance: The robot avoids obstacles earlier in advance based on the currently optimized reference path, especially in narrow roads or with multiple obstacles. A reference path is generated that is as far away from obstacles as possible to improve the robot's obstacle avoidance ability and avoid collision accidents.

[0040] 5. Multi-sensor fusion: In order to perceive the environment more accurately, the robot uses multiple sensors for data fusion, including but not limited to lidar, cameras, ultrasonic sensors, etc., so as to obtain more comprehensive and accurate real-time environmental information and improve the effects of path planning and obstacle avoidance.

[0041] 6. Adaptive motion control: Based on the real-time environmental information and path planning results, the robot can adaptively adjust the motion strategy and control parameters, including speed, acceleration, steering angle, etc., to achieve more flexible and efficient motion control and improve the adaptability and robustness of the robot.

[0042] This embodiment uses A* and DWA algorithms to realize path generation and obstacle avoidance of wheeled mobile robots. Specifically:

[0043] Step 1: In the local field of view within a few meters in the direction of the robot's advance, use sensors such as laser radar and cameras to sense and collect environmental information in real time. In this embodiment, the local field of view is set to 5 meters, and the robot will collect environmental information within 5 meters in front, including the location, shape, distance, etc. of obstacles.

[0044] Step 2: Based on the observed real-time environment within the local field of view, the global reference path is updated at a fixed frequency to generate a better local reference path suitable for the current environment. In this embodiment, the global reference path is updated once a second. Then, every second, the robot will recalculate and update the global reference path based on the latest environmental information. During the path planning process, the robot's motion constraints, such as maximum speed, minimum turning radius, etc., are always considered to ensure that the generated path is feasible and suitable for the robot to execute. In this embodiment, the maximum speed of the robot is set to 0.5 meters per second and the minimum turning radius is 5 meters. The generated path will meet the robot's motion constraints, ensuring that the robot will not collide with obstacles during navigation and obstacle avoidance.

[0045] Step 3: Use the DWA local path algorithm to generate a local reference path based on the global reference path and real-time environment information. In this embodiment, the DWA algorithm generates a local reference path from the current position to the target point based on the global reference path and real-time environment information.

[0046] Step 4: Based on the optimized reference path, the robot can avoid obstacles earlier and generate a reference path that is as far away from obstacles as possible in a narrow or multi-obstacle environment. In this embodiment, the maximum speed of the robot is set to 0.5 m / s and the minimum turning radius is 5 m. The generated reference path will meet the robot's motion constraints, ensuring that the robot will not collide with obstacles during navigation and obstacle avoidance.

[0047] Through the above steps, this embodiment can effectively solve the problems of static global path, insufficient local path generation and inaccurate local global path update in the prior art, and improve the robot's navigation and obstacle avoidance capabilities in complex environments.

[0048] In summary, compared with the prior art, this solution can solve the following technical problems:

[0049] 1. The static problem of global path: In the existing technical solutions, the global path is generated based on the static environment. When the environment changes, the global path may no longer be applicable and needs to be replanned, which will cause certain delays and resource consumption. This solution solves the static problem of the global path by updating the global reference path at a fixed frequency within the local field of view according to the observed real-time environment, so as to generate a better local reference path suitable for the current environment.

[0050] 2. Insufficient local path generation: Although existing local path generation algorithms, such as DWA, can be adjusted according to the real-time environment, the generated path may still not be ideal in complex environments or narrow spaces and cannot completely avoid obstacles, affecting the safety and efficiency of the robot. This solution selects appropriate target points during the local fixed-frequency path search process, and the generated path meets the robot's motion constraints and stays away from nearby obstacles, thereby solving the problem of insufficient local path generation.

[0051] 3. Inaccurate update of local global path: When updating the local global path, the existing technical solutions may not be able to obtain the latest environmental information in a timely and accurate manner, resulting in inaccurate local reference paths, which affects the robot's navigation and obstacle avoidance effects. This solution is designed to update the global reference path at a fixed frequency within the local field of view according to the observed real-time environment, so as to generate a better local reference path suitable for the current environment, thereby solving the problem of inaccurate update of the local global path.

Claims

1. A method for dynamically optimizing and generating a robot path, characterized in that: The following steps are involved: S1. Obtain real-time environmental observation information within the robot's local field of view; S2, based on the global reference path, combined with real-time environmental observation information and robot motion constraints, the global reference path is updated at a fixed frequency; S3. Based on the real-time environmental observation information and the updated global reference path, the DWA local path algorithm is used to generate a local reference path.

2. A robot path dynamic optimization generation method according to claim 1, characterized in that: The step S1 specifically involves acquiring real-time environmental observation information through sensors carried by the robot.

3. A robot path dynamic optimization generation method according to claim 2, characterized in that: The sensors include laser radar, depth camera, ultrasonic radar, and obstacle avoidance radar.

4. A robot path dynamic optimization generation method according to claim 2, characterized in that: The real-time environmental observation information is specifically environmental information within a set distance range in front of the robot.

5. A robot path dynamic optimization generation method according to claim 4, characterized in that: The environmental information includes the location, shape, and distance information of obstacles.

6. A method for dynamically optimizing and generating a robot path according to claim 5, characterized in that: The global reference path in step S2 is specifically based on a pre-constructed map, according to obstacle information on the map, and through a path search algorithm, the path coordinates from the current position to the target position are planned on the grid map.

7. A robot path dynamic optimization generation method according to claim 1, characterized in that: The robot motion constraints in step S2 include maximum speed and minimum turning radius.

8. A robot path dynamic optimization generation method according to claim 1, characterized in that: The fixed-frequency updating of the global reference path in step S2 specifically involves regenerating the global reference path using the A* algorithm at a preset time interval.

9. A method for dynamically optimizing and generating a robot path according to claim 6, characterized in that: The step S3 specifically selects a path point on the global reference path within the current visible field of view as the target point of the current travel, and uses the DWA local path algorithm to generate a local reference path from the current position to the target point.

10. A robot path dynamic optimization generation method according to claim 9, characterized in that: The target point of the current movement is adjusted according to the real-time environmental observation information. If there is no obstacle within the set distance range in front of the robot, a path point closest to the current position is directly selected from the global reference path; If there is an obstacle within a set distance in front of the robot, a path point is selected from the global reference path whose distance to the obstacle meets a preset threshold.

Images