ROS2-based routing inspection robot mapping and multi-point navigation method

Abstract

The invention discloses an ROS2-based routing inspection robot mapping and multi-point navigation method, which introduces an advanced Cartographic algorithm for mapping, combines an A * algorithm and a TEB algorithm for path planning, and realizes multi-point navigation, so that a robot can more quickly and flexibly deal with a complex environment and meet diversified routing inspection task requirements. Meanwhile, perfect system configuration and environment construction are adopted, including a Ubuntu system of a specific version, an ROS 2 issuing version and collaborative configuration of related software and hardware, and guarantee is provided for stable operation of the whole system.

Description

Technical Field

[0001] The present invention relates to the field of autonomous positioning and navigation of mobile robots, and in particular to a mapping and multi-point navigation method of an inspection robot based on ROS2. Background Art

[0002] In the field of inspection robots, common technologies include the use of basic sensors for environmental perception, which has limitations. Existing inspection robots often perform tasks according to predetermined routes or remote control, lack the ability to flexibly adjust to the actual equipment conditions, and the fixed routes are easily restricted by the working environment and prone to collisions. The inflexible characteristics make them inconvenient to use, and they are not easy to navigate autonomously and operate stably in complex environments, resulting in low efficiency.

[0003] Existing technical solutions usually obtain environmental information through sensors, and then build maps and plan paths based on preset rules and algorithms. Some existing inspection robot systems use traditional mapping algorithms, such as grid-based map building methods. In terms of path planning, a relatively simple global path planning algorithm, such as the Dijkstra algorithm, may be used, while local path planning may rely on simple obstacle avoidance rules.

[0004] However, the existing inspection robot system mapping algorithm is not advanced enough and has difficulty in processing complex environments and large amounts of data, resulting in insufficient map accuracy and completeness. When implementing multi-point navigation, fixed routes are easily restricted by the working environment and prone to collisions, making it difficult to meet the inspection task requirements in complex environments. In addition, the efficiency and adaptability of the path planning algorithm are poor, and it may not be able to find the optimal path or respond to dynamic obstacles in a timely manner in complex environments. Summary of the invention

[0005] The purpose of the present invention is to solve the problems in the prior art and provide a mapping and multi-point navigation method for an inspection robot based on ROS2, introduce the advanced Cartographer algorithm for mapping, combine the A* algorithm and TEB algorithm for path planning, and realize multi-point navigation, so that the robot can respond to complex environments more quickly and flexibly and meet the diverse inspection task requirements.

[0006] In order to achieve the above object, the technical solution of the present invention is:

[0007] A method for mapping and multi-point navigation of inspection robots based on ROS2. Step 1: configure the Ubuntu 22.04 system, ROS distribution version humble and system environment required by the system; use Gazebo software to build a three-dimensional simulation world model, establish a working environment, and write configuration files to build an actual mobile robot platform, multi-line laser radar and depth camera for experimental simulation;

[0008] Step 2: Download the Cartographer algorithm function packages ros-humble-cartographer and ros-humble-cartographer-ros, install the Cartographer algorithm into the working environment created in step 1 and compile the algorithm function package;

[0009] Step 3: Download, install and compile the ros-humble-navigation2 package, establish a complete working environment, and configure relevant parameters;

[0010] Step 4: Use the Cartographer algorithm to build a ground map, run the multi-line laser radar driver and the laser point cloud 3D harness to convert into a 2D harness program to more fully extract the surrounding environment features, configure the topic name and function package parameters, run the mobile robot chassis driver and make the posture conversion relationship between the chassis and the radar, run the Cartographer program and drive the chassis to build a 2D map;

[0011] Step 5: Select and configure the positioning algorithm, global path planning algorithm and local path planning algorithm;

[0012] Step 6: Set up a multi-point navigation task, obtain the coordinates of multiple inspection target points by calling the nav2_util function package in ROS2, import the developed control function package, and execute the function package to control the robot model to reach multiple target points in sequence.

[0013] Preferably, step two also includes: processing the measured radar point cloud data, comparing it with the sub-map obtained in the previous scan, and if the match is successful, inserting it into the previous sub-map for updating and optimization, and generating the next sub-map; as the robot moves in the target area, repeatedly obtaining the sub-map; when no new sub-map is inserted, the construction of the complete sub-map is completed.

[0014] Preferably, after the construction of multiple sub-graphs is completed, the back-end starts loop detection. After each sub-graph is constructed, it will be matched with all previous motion-filtered point cloud frames to perform loop optimization and eliminate accumulated errors.

[0015] Preferably, step three also includes: Nav2 navigation is internally divided into a BT navigation server, and according to the behavior tree description file in XML format, the following three server modules are configured and called to complete the behavior control of the robot:

[0016] The planner server is responsible for global path planning and planning the path from the starting point to the target point;

[0017] The control server completes the real-time control of the robot based on the global path, combined with real-time obstacles and local cost maps;

[0018] The recovery server is responsible for performing recovery operations such as turning and retreating when the robot encounters special situations such as path blockage and jamming, and notifies the planning server to update the global path.

[0019] Preferably, step five also includes: configuring the Cartographer algorithm parameter file to provide a positioning function, loading the previously constructed two-dimensional pbstream format map; selecting the A* algorithm for the global path planning algorithm, configuring the TEB algorithm for the local path planning, and adjusting the cost map and path planning parameter file.

[0020] Preferably, step six also includes: issuing a prompt voice when reaching the target point.

[0021] The beneficial effects of the present invention are:

[0022] 1. Use complete system configuration and environment construction, including a specific version of Ubuntu system, ROS 2 distribution, and coordinated configuration of related software and hardware to ensure the stable operation of the entire system.

[0023] 2. The advanced Cartographer algorithm is used for map construction. Its efficient data processing and sub-map comparison, update and loop detection mechanism improves the accuracy and integrity of the map and can more accurately reflect the complex inspection environment. The existing mapping technology has low accuracy and it is difficult to present environmental details.

[0024] 3. The A* algorithm and TEB algorithm are integrated for path planning. The A* algorithm ensures the optimal solution for the global path, and the TEB algorithm ensures dynamic obstacle avoidance and efficient calculation of the local path. The combination of the two improves the overall performance of path planning.

[0025] 4. The multi-point navigation function is realized. Through a specific control package and coordinate acquisition method, the robot can accurately reach multiple inspection target points in sequence to meet more complex and diverse inspection task requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 is a flow chart of the method of the present invention;

[0027] Figure 2 This is the Nav2 navigation schematic. DETAILED DESCRIPTION

[0028] In order to enable those skilled in the art to better understand the scheme of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and implementation methods.

[0029] like Figure 1-2 As shown, a method for mapping and multi-point navigation of an inspection robot based on ROS2 includes the following steps:

[0030] Step 1: Configure the Ubuntu 22.04 system, ROS distribution version humble and system environment required by this system. Use Gazebo software to build a 3D simulation world model, establish a working environment, and write configuration files to build an actual mobile robot platform, multi-line laser radar and depth camera for experimental simulation.

[0031] Step 2: Download the Cartographer algorithm function package ros-humble-cartographer and ros-humble-cartographer-ros, install the cartographer algorithm into the working environment established in step 1 and compile the algorithm function package. The Cartographer algorithm works as follows: the measured radar point cloud data is processed and compared with the sub-image obtained in the previous scan. If the match is successful, it is inserted into the previous sub-image for update and optimization, and the next sub-image is generated. As the robot moves in the target area, the sub-image is repeatedly obtained. When no new sub-image is inserted, the construction of the complete sub-image is completed. After the construction of multiple sub-images is completed, the back-end will start loop detection. After each sub-image is built, it will be matched with all the previous motion-filtered point cloud frames for loop optimization to eliminate the accumulated error.

[0032] Step 3: Download, install and compile the ros-humble-navigation2 package, establish a complete working environment, and configure relevant parameters. Nav2 navigation is internally divided into BT navigation servers. According to the behavior tree description file in XML format, the following three server modules are configured to complete the behavior control of the robot: the planner server is responsible for global path planning and is responsible for planning the path from the starting point to the target point; the control server completes the real-time control of the robot based on the global path, combined with real-time obstacles and local cost maps; the recovery server is responsible for turning, retreating and other recovery operations when the robot encounters special obstacles such as path blockage and jamming, and notifies the planning server to update the global path.

[0033] Step 4: Use the Cartographer algorithm to create a ground map, run the multi-line lidar driver and the laser point cloud 3D harness to convert into a 2D harness program to more fully extract the surrounding environment features, configure the topic name and function package parameters, run the mobile robot chassis driver and make the posture conversion relationship between the chassis and the radar, run the Cartographer program and drive the chassis to build a 2D map.

[0034] Step 5: Select and configure the positioning algorithm, global path planning algorithm, and local path planning algorithm. Configure the Cartographer algorithm parameter file to provide positioning function, load the previously built two-dimensional pbstream format map; select the A* algorithm for the global path planning algorithm, which can ensure the optimal solution, avoid useless searches, and has strong adaptability; configure the TEB algorithm for local path planning, which has good dynamic obstacle avoidance capabilities, a wide range of applications, and high computational efficiency; adjust the cost map and path planning parameter file.

[0035] Step 6: Set up a multi-point navigation task. Get the coordinates of multiple inspection target points by calling the nav2_util function package in ROS2, import the developed control function package, execute the function package to control the robot model to reach multiple target points in sequence, and issue a prompt voice when reaching the target point.

[0036] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A ROS2-based inspection robot mapping and multi-point navigation method, characterized by: The following steps are included: Step 1: Configure the Ubuntu 22.04 system, ROS distribution version humble and system environment required by the system; use Gazebo software to build a 3D simulation world model, establish a working environment, and write configuration files to build an actual mobile robot platform, multi-line laser radar and depth camera for experimental simulation; Step 2: Download the Cartographer algorithm function packages ros-humble-cartographer and ros-humble-cartographer-ros, install the Cartographer algorithm into the working environment created in step 1 and compile the algorithm function package; Step 3: Download, install and compile the ros-humble-navigation2 package, establish a complete working environment, and configure relevant parameters; Step 4: Use the Cartographer algorithm to build a ground map, run the multi-line laser radar driver and the laser point cloud 3D harness to convert into a 2D harness program to more fully extract the surrounding environment features, configure the topic name and function package parameters, run the mobile robot chassis driver and make the posture conversion relationship between the chassis and the radar, run the Cartographer program and drive the chassis to build a 2D map; Step 5: Select and configure the positioning algorithm, global path planning algorithm and local path planning algorithm; Step 6: Set up a multi-point navigation task, obtain the coordinates of multiple inspection target points by calling the nav2_util function package in ROS2, import the developed control function package, and execute the function package to control the robot model to reach multiple target points in sequence.

2. The ROS2-based inspection robot mapping and multi-point navigation method according to claim 1 is characterized in that: Step 2 also includes: processing the measured radar point cloud data, comparing it with the sub-map obtained in the previous scan, inserting it into the previous sub-map for updating and optimization if the match is successful, and generating the next sub-map; as the robot moves in the target area, the sub-map is repeatedly obtained; when no new sub-map is inserted, the construction of the complete sub-map is completed.

3. The ROS2-based inspection robot mapping and multi-point navigation method according to claim 2 is characterized in that: After the construction of multiple sub-images is completed, the backend starts loop detection. After each sub-image is constructed, it will be matched with all the previous motion-filtered point cloud frames to perform loop optimization and eliminate accumulated errors.

4. The ROS2-based inspection robot mapping and multi-point navigation method according to claim 1 is characterized in that: Step 3 also includes: Nav2 navigation is divided into BT navigation server. According to the behavior tree description file in XML format, the following three server modules are configured and called to complete the behavior control of the robot: The planner server is responsible for global path planning and planning the path from the starting point to the target point; The control server completes the real-time control of the robot based on the global path, combined with real-time obstacles and local cost maps; The recovery server is responsible for performing recovery operations when the robot encounters path blockage, jamming, etc., and notifying the planning server to update the global path.

5. The ROS2-based inspection robot mapping and multi-point navigation method according to claim 1 is characterized in that: Step five also includes: configuring the Cartographer algorithm parameter file to provide positioning function, loading the previously built two-dimensional pbstream format map; selecting the A* algorithm for the global path planning algorithm, configuring the TEB algorithm for the local path planning, and adjusting the cost map and path planning parameter file.

6. The ROS2-based inspection robot mapping and multi-point navigation method according to claim 1 is characterized in that: Step six also includes: issuing a prompt voice when reaching the target point.

Images