An autonomous navigation robot based on ROS

Abstract

The present invention discloses an autonomous navigation robot based on ROS, and relates to the field of autonomous navigation robots of ROS, including robot system kinematic modeling and model simulation, mobile robot SLAM and navigation algorithm, ROS Mecanum wheel mobile robot platform construction, robot SLAM and positioning, and path planning experimental analysis, wherein the robot system kinematic modeling and model simulation are composed of robot coordinate system establishment, robot laser radar observation model, Mecanum wheel kinematic model analysis, environmental map model, robot model and motion environment simulation, and the mobile robot SLAM and navigation algorithm are composed of mobile robot SLAM and mobile robot navigation algorithm research. The present invention mathematically models a mobile robot based on a Mecanum wheel, including establishing a coordinate system, a laser radar observation model, calculating the forward and inverse solutions of the Mecanum wheel kinematic model, and establishing a robot URDF model and a Gazebo space motion environment.

Description

Technical Field

[0001] The present invention relates to the field of autonomous navigation robots of ROS, and in particular to an autonomous navigation robot based on ROS. Background Art

[0002] With the rapid development of the robotics industry in recent years, autonomous mobile robots have become a hot spot for development and investment in the robotics industry, and are also the main trend in the development of robotics technology. Autonomous mobile robots can perceive the surrounding unfamiliar environment and estimate their own position and status through sensors, and realize autonomous movement from the starting point to the target point in an unfamiliar environment with obstacles, thereby meeting people's operational needs. Autonomous mobile robots are widely used in military, industrial and agricultural, medical and health fields due to their advantages such as small size, flexibility and simple operation, such as material transportation, warehouse inspection, epidemic prevention and disinfection, etc.

[0003] However, the application of non-omnidirectional wheeled mobile robots in many fields is limited due to their shortcomings such as large turning radius and inflexible movement. Summary of the invention

[0004] The present invention provides the following technical scheme: an autonomous navigation robot based on ROS, comprising robot system kinematic modeling and model simulation, mobile robot SLAM and navigation algorithm, ROS Mecanum wheel mobile robot platform construction, robot SLAM and positioning and path planning experimental analysis, wherein the robot system kinematic modeling and model simulation are composed of robot coordinate system establishment, robot laser radar observation model, Mecanum wheel kinematic model analysis, environmental map model, robot model and motion environment simulation, the mobile robot SLAM and navigation algorithm are composed of mobile robot SLAM, mobile robot navigation algorithm research, the ROS Mecanum wheel mobile robot platform construction is composed of mobile robot hardware platform construction, robot host computer system, ROS Mecanum wheel mobile robot platform, the robot SLAM and positioning and path planning experimental analysis are composed of robot motion test environment, robot SLAM experiment, AMCL indoor positioning experiment, path planning navigation experiment.

[0005] Preferably, the robot coordinate system is established with three degrees of freedom, namely, translation of the plane where the robot is located and rotation around an axis perpendicular to the plane where the robot is located. The coordinate system WWWYOX is an inertial coordinate system, RRRxoy is a robot body coordinate system, the origin Ro is the geometric center of the robot, the Rx axis is the positive direction of the robot movement, and the robot's position and posture can be represented by a vector ξ composed of elements x, y, wherein x and y represent the plane position of the robot in the inertial coordinate system, and is the robot attitude angle, which is expressed as the angle between the Rx axis of the robot coordinate system and the XW axis of the inertial coordinate system, and is used to describe the movement direction of the robot;

[0006] The robot laser radar observation model is obtained through the working principle of triangulation laser radar, the ranging principle of TOF laser radar, and the performance analysis of triangulation laser radar and TOF laser radar. The relationship between the environmental feature position observed by the laser radar and the robot posture coordinates in the inertial coordinate system is usually expressed by a nonlinear discrete time difference equation.

[0007] The kinematic model analysis of the Mecanum wheel has four Mecanum wheels with the same orientation, the vertical distance between the horizontal center axis of wheel l and the Rx axis is a, and the vertical distance between the longitudinal center axis of wheel l and the Ry axis is b;

[0008] The environmental map models include grid maps, topological maps, feature maps and hybrid maps. The robot model and the robot model in the motion environment simulation use URDF to establish a robot simulation model. The URDF file uses the extensible markup language XML to describe the robot's rigid body appearance, coordinate position, joint type and other aspects, and presents them in the Rviz interface. At the same time, it can also display sensor data, robot motion status, changes in the surrounding environment and other information in real time in a graphical manner. The robot's joint status is updated by receiving speed control instructions, thereby realizing the movement of the robot in Rviz. The motion environment simulation in the robot model and the motion environment simulation uses the Gazebo simulator to design the robot motion environment. The robot's environment is a rectangular enclosed space, which is divided into a corridor and three small rooms. The rooms are equipped with bookshelves, tables, ladders and other items. The robot can move freely in the environment through the small doors between the corridor and each room.

[0009] Preferably, the SLAM of the mobile robot is composed of a classic SLAM method architecture, a Rao-Blackwellised particle filter algorithm, a gmapping algorithm, and a gmapping algorithm mapping simulation. The classic SLAM method architecture uses a filter to perform state estimation on the collected environmental data and the measured robot motion model, and then obtains map data and robot posture. The Rao-Blackwellised particle filter algorithm regards the system state and measurement information as random variables, which conforms to the actual situation of mobile robot SLAM and is suitable for nonlinear, non-Gaussian mobile robot systems, avoiding any approximate linearization processing. The gmapping algorithm can simultaneously perform robot positioning and mapping by subscribing to sensor information and completing some necessary parameter configurations. The gmapping algorithm mapping simulation collects radar data of the unfamiliar environment around the robot to generate a grid map.

[0010] Preferably, the mobile robot navigation algorithm research is composed of a multi-sensor fusion algorithm based on an extended Kalman filter, an ACML positioning algorithm, and a mobile robot path planning algorithm;

[0011] The adaptive Monte Carlo positioning algorithm in the multi-sensor fusion algorithm of the extended Kalman filter obtains the optimal final strategy through continuous strategy iteration. The characteristic of the MC control algorithm is to obtain the phased task according to the current strategy, and then update the value through all states in the multi-task, and then reach the convergence of the average value of the state value function through iteration. The AMCL positioning algorithm of ROS in the multi-sensor fusion algorithm of the extended Kalman filter is used for the global positioning of the robot in the map;

[0012] The ACML positioning algorithm is a topic subscribed and published in ROS, and the AMCL algorithm simulation in the multi-sensor fusion algorithm of the extended Kalman filter is a robot positioning simulation and a posture tracking simulation;

[0013] The mobile robot path planning algorithm is composed of the A* global path planning algorithm, the move_base navigation framework, and the robot path planning simulation;

[0014] The mobile robot hardware platform is built to verify and analyze the robot's SLAM, path planning, navigation and other algorithms, and uses the Raspberry Pi version 4B, in which the Ubuntu 18.04 operating system is installed and the ROS framework is encapsulated in the system, and the function package required for the experiment is installed as the top-level control platform of the robot, in which the robot's SLAM, positioning, and path planning are completed, and TM32F405RGT6 is used as the MCU of the underlying control board. The functions used include general I / O, general timer, general serial port, timer orthogonal decoding, timer PWM, IIC bus, and ADC voltage sampling;

[0015] The ROS Mecanum wheel mobile robot platform is a robot, and the length, width and height of the robot are 30 cm, 24 cm and 18 cm respectively.

[0016] Preferably, the robot SLAM experiment places the robot at an initial position, controls the robot to move in the room, and enables the laser radar to scan the entire room. After the scanning is completed, a map is generated. It can be seen from the figure that the robot can generate a complete room map. The lines of the room walls shown in the map are smooth, the details of the internal environment are clear, and the constructed map environment is highly consistent with the actual environment.

[0017] The AMCL indoor positioning experiment uses the AMCL algorithm to position the robot. The robot positioning process, where the red arrow is the particle swarm generated by the algorithm, and the arrow direction indicates the posture of the robot; the red dot is the boundary position of the object displayed by the laser radar;

[0018] The path planning and navigation experiment uses the move_base navigation algorithm to perform a path planning experiment on the robot, and experimentally analyzes the global path planning and local path planning of the robot in an actual scenario.

[0019] Technical effects and advantages of the present invention:

[0020] 1. The present invention performs mathematical modeling on a Mecanum wheel-based mobile robot, including establishing a coordinate system, a laser radar observation model, calculating the forward and inverse solutions of the Mecanum wheel kinematic model, and establishing a robot URDF model and a Gazebo space motion environment.

[0021] 2. The present invention studies the SLAM, global positioning and path planning algorithms of mobile robots. In the SLAM experiment, the gmapping algorithm based on the RBPF algorithm is selected for map construction. The simulation results show that the map can be constructed in real time with high accuracy in small indoor scenes. The multi-sensor fusion algorithm based on the extended Kalman filter is studied, and the encoder data and IMU data are integrated to obtain more accurate odometer data. The odometer data is applied to the AMCL positioning algorithm to improve the robot positioning accuracy.

[0022] 3. The present invention conducts an overall design of the mobile robot, and builds a Mecanum wheel omnidirectional robot mobile platform based on hardware modules such as Raspberry Pi, STM32F405, IMU, lidar and DC motor. The algorithm studied in this paper is configured in the robot, and robot SLAM, positioning and path planning experiments are carried out. The experimental results show that the map construction effect of the robot designed in this paper is basically consistent with the actual map. The AMCL algorithm can accurately determine the robot's position and posture. The move_base navigation framework that integrates the A* and DWA algorithms can effectively complete path planning and navigation obstacle avoidance experiments, indicating that the autonomous navigation robot designed in this paper has high adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 It is a schematic diagram of the inertial coordinate system and the robot local coordinate system of the present invention;

[0024] Figure 2 A schematic diagram of the relationship between the motion nodes of the robot of the present invention;

[0025] Figure 3 It is a schematic diagram of the gmapping algorithm flow of the present invention;

[0026] Figure 4 It is a schematic diagram of the basic flow of the packet algorithm of the present invention;

[0027] Figure 5 A schematic diagram of the AMCL positioning process of the robot of the present invention during movement;

[0028] Figure 6 It is a block diagram of the mobile robot platform of the present invention. DETAILED DESCRIPTION

[0029] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Specific embodiment one:

[0031] like Figure 1 As shown, the robot system kinematic modeling and model simulation consists of the establishment of the robot coordinate system, the robot lidar observation model, the Mecanum wheel kinematic model analysis, the environmental map model, the robot model and the motion environment simulation. The mobile robot SLAM and navigation algorithm consists of the mobile robot SLAM and mobile robot navigation algorithm research. The ROS Mecanum wheel mobile robot platform construction consists of the mobile robot hardware platform construction, the robot host computer system, and the ROS Mecanum wheel mobile robot platform. The robot SLAM and positioning and path planning experimental analysis consists of the robot motion test environment, the robot SLAM experiment, the AMCL indoor positioning experiment, and the path planning navigation experiment.

[0032] like Figure 1-2 As shown, the robot coordinate system is established with three degrees of freedom, which are the translation of the robot's plane and the rotation around the axis perpendicular to the robot's plane. The coordinate system WWWYOX is the inertial coordinate system, RRRxoy is the robot body coordinate system, the origin Ro is the robot's geometric center, the Rx axis is the positive direction of the robot's movement, and the robot's position and posture can be represented by a vector ξ composed of elements x, y, where x and y represent the robot's plane position in the inertial coordinate system, and is the robot's attitude angle, which is expressed as the angle between the Rx axis of the robot coordinate system and the XW axis of the inertial coordinate system, and is used to describe the robot's movement direction;

[0033] like Figure 1-2As shown in the figure, the mobile robot system is modeled, and the kinematic model and robot simulation model of the omnidirectional mobile robot are analyzed and established. The inertial coordinate system and the robot coordinate system are established; the working principle of the commonly used laser radar is analyzed, the triangulation laser radar is selected and the laser radar observation model is established, the kinematic model analysis of the omnidirectional wheel is completed, the kinematic model of the Mecanum wheel robot is established, and the forward and inverse solutions of the Mecanum wheel kinematic model are given. The characteristics of different environmental maps are analyzed, and the grid map is selected as the map for subsequent research according to needs. The mobile robot model is written in XML language, the overall structure diagram of the robot model is given and visualized in Rviz, and Gazebo is used to establish a rectangular closed simulation space as the robot's motion environment. The robot model and simulation space environment will be used in subsequent algorithm research and analysis. Specific embodiment 2:

[0035] like Figure 2-3 As shown in the figure, the robot laser radar observation model is obtained through the working principle of triangulation laser radar, the ranging principle of TOF laser radar, and the performance analysis of triangulation laser radar and TOF laser radar. The relationship between the environmental feature position observed by the laser radar and the robot posture coordinates in the inertial coordinate system is usually expressed by a nonlinear discrete time difference equation. The Mecanum wheel kinematic model analysis has 4 Mecanum wheels with the same orientation. The vertical distance between the horizontal center axis of wheel l and the Rx axis is a, and the vertical distance between the longitudinal center axis of wheel l and the Ry axis is b.

[0036] like Figure 2-3 As shown, the environmental map models include grid maps, topological maps, feature maps and hybrid maps. The robot model in the robot model and motion environment simulation uses URDF to establish the robot simulation model. The URDF file uses the extensible markup language XML to describe the robot's rigid body appearance, coordinate position, joint type and other aspects, and presents them in the Rviz interface. At the same time, it can also display sensor data, robot motion status, changes in the surrounding environment and other information in real time in a graphical way. By receiving speed control instructions, the robot's joint status is updated to achieve the movement of the robot in Rviz. The motion environment simulation in the robot model and motion environment simulation uses the Gazebo simulator to design the robot's motion environment. The robot's environment is a rectangular enclosed space, which is divided into a corridor and three small rooms. There are bookshelves, tables, ladders and other items in the room. The robot can move freely in the environment through the corridor and the small doors between the rooms.

[0037] like Figure 2-3As shown in the figure, the gmapping algorithm based on RBPF is analyzed, and the real-time positioning and map construction of the robot simulation environment are realized based on the algorithm. Then, the multi-sensor fusion algorithm based on the extended Kalman filter is studied, and the encoder data and IMU inertial unit data are integrated using the EKF package provided in ROS to obtain more accurate odometer data; then the AMCL positioning algorithm is analyzed, and the simulation results show that the algorithm can better determine the robot's posture when performing robot positioning indoors, and can also show a relatively ideal effect in posture tracking simulation; finally, the principles of the A* global path planning algorithm and the DWA local path planning algorithm are studied, and the move_base navigation framework that integrates the A* algorithm and the DWA algorithm is used for path planning simulation. The simulation results show that the move_base navigation framework can better realize the robot's path planning, navigation and obstacle avoidance functions. Specific embodiment three:

[0039] like Figure 3-4 As shown, the research on mobile robot navigation algorithm is composed of multi-sensor fusion algorithm and ACML positioning algorithm based on extended Kalman filter and mobile robot path planning algorithm. The adaptive Monte Carlo positioning algorithm in the multi-sensor fusion algorithm of extended Kalman filter obtains the optimal final strategy through continuous strategy iteration. The characteristic of MC control algorithm is to obtain phased tasks according to the current strategy, and then update the value through all states in multi-tasks, and then reach the convergence of the average value of state value function through iteration. The AMCL positioning algorithm of ROS in the multi-sensor fusion algorithm of extended Kalman filter is used for the global positioning of the robot in the map.

[0040] like Figure 2-4 As shown in the figure, the topics subscribed and published by the ACML positioning algorithm in ROS, the AMCL algorithm simulation in the multi-sensor fusion algorithm of the extended Kalman filter is the robot positioning simulation and posture tracking simulation, and the mobile robot path planning algorithm is composed of the A* global path planning algorithm, the move_base navigation framework, and the robot path planning simulation;

[0041] like Figure 2-3 As shown in the figure, the construction of the mobile robot hardware platform is to verify and analyze the robot's SLAM, path planning, navigation and other algorithms, and the Raspberry Pi version 4B is used, in which the Ubuntu 18.04 operating system is installed and the ROS framework is encapsulated in the system, and the function package required for the experiment is installed. As the top-level control platform of the robot, the robot's SLAM, positioning, and path planning are completed in it, and TM32F405RGT6 is used as the MCU of the bottom control board. The functions used include general I / O, general timer, general serial port, timer orthogonal decoding, timer PWM, IIC bus, and ADC voltage sampling;

[0042] like Figure 2-5 As shown in the figure, the Mecanum wheel mobile robot platform of ROS is a robot with a length, width and height of 30cm, 24cm and 18cm respectively. A Mecanum wheel robot based on ROS is built above. First, the overall design of the robot platform is carried out. Secondly, Raspberry Pi 4B, STM32F405, Silan RPLIDARA1 laser radar, MPU6050, TB6612FNG motor driver chip, JGB37-520 brushed DC motor, incremental encoder, Mecanum wheel and other hardware modules are selected to build the robot platform. Then, Ubuntu18.04 system and ROS robot framework are installed in Raspberry Pi, and relevant files are configured; finally, PID parameters of the robot are tuned. Specific embodiment four:

[0044] like Figure 4-6 As shown in the figure, the robot SLAM experiment places the robot at the initial position, controls the robot to move in the room, and makes the laser radar scan the entire room. After the scanning is completed, the generated map shows that the robot can generate a complete room map. The lines of the room walls shown in the map are smooth, the details of the internal environment are clear, and the constructed map environment is highly consistent with the actual environment.

[0045] like Figure 4-6 As shown in the figure, the AMCL indoor positioning experiment uses the AMCL algorithm to locate the robot. The robot positioning process, where the red arrow is the particle swarm generated by the algorithm, the arrow direction indicates the robot's posture, and the red dot is the object boundary position displayed by the lidar;

[0046] like Figure 4-6 As shown in the figure, the path planning and navigation experiment uses the move_base navigation algorithm to perform a path planning experiment on the robot, and conducts experimental analysis on the global path planning and local path planning of the robot in actual scenarios;

[0047] like Figure 4-6As shown in the figure, firstly, the gmapping algorithm is used to conduct SLAM experiments in the laboratory environment to establish a map. The experimental results show that the established map is consistent with the real scene, the environmental details are relatively clear, and the map data can be used in subsequent experiments. Then the AMCL algorithm is used to position the robot. In the robot positioning experiment and posture tracking experiment, the algorithm can accurately determine the robot's posture. Finally, the robot path planning experiment is carried out. In the original map environment, the robot can successfully complete the navigation experiment. When a fixed obstacle appears on the forward path, the robot can replan the path and effectively avoid the obstacle. When the robot encounters a pedestrian, it can adjust the planned path in real time according to the pedestrian's movement and bypass the pedestrian. The experimental effect is relatively ideal. The robot can successfully realize functions such as navigation and obstacle avoidance under different circumstances.

[0048] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. An autonomous navigation robot based on ROS, including robot system kinematic modeling and model simulation, mobile robot SLAM and navigation algorithm, ROS Mecanum wheel mobile robot platform construction, robot SLAM and positioning and path planning experimental analysis, characterized by: The robot system kinematic modeling and model simulation consists of robot coordinate system establishment, robot lidar observation model, Mecanum wheel kinematic model analysis, environment map model, robot model and motion environment simulation; the mobile robot SLAM and navigation algorithm consists of mobile robot SLAM and mobile robot navigation algorithm research; the ROS Mecanum wheel mobile robot platform construction consists of mobile robot hardware platform construction, robot host computer system, and ROS Mecanum wheel mobile robot platform; the robot SLAM and positioning and path planning experimental analysis consists of robot motion test environment, robot SLAM experiment, AMCL indoor positioning experiment, and path planning navigation experiment; The robot coordinate system is established with three degrees of freedom, which are the translation of the plane where the robot is located and the rotation around the axis perpendicular to the plane where the robot is located. The coordinate system WWWYOX is the inertial coordinate system, RRRxoy is the robot body coordinate system, the origin Ro is the robot geometric center, the Rx axis is the positive direction of the robot movement, and the robot's posture is represented by a vector ξ composed of elements x, y, where x and y represent the plane position of the robot in the inertial coordinate system, and is the robot posture angle, which is expressed as the angle between the Rx axis of the robot coordinate system and the XW axis of the inertial coordinate system, and is used to describe the movement direction of the robot; The robot laser radar observation model is obtained through the working principle of triangulation laser radar, the ranging principle of TOF laser radar, and the performance analysis of triangulation laser radar and TOF laser radar. The relationship between the environmental feature position observed by the laser radar and the robot posture coordinates in the inertial coordinate system is expressed by a nonlinear discrete time difference equation. The kinematic model analysis of the Mecanum wheel has four Mecanum wheels with the same orientation, the vertical distance between the horizontal center axis of wheel l and the Rx axis is a, and the vertical distance between the longitudinal center axis of wheel l and the Ry axis is b; The environmental map model includes a grid map, a topological map, a feature map and a hybrid map. The robot model and the robot model in the motion environment simulation use URDF to establish a robot simulation model. The URDF file uses the extensible markup language XML to describe the rigid body appearance, coordinate position, and joint type of the robot, and presents it in the Rviz interface. At the same time, it displays sensor data, robot motion status, and surrounding environment change information in real time in a graphical manner. The robot's joint status is updated by receiving speed control instructions, thereby realizing the movement of the robot in Rviz. The motion environment simulation in the robot model and the motion environment simulation uses the Gazebo simulator to design the robot motion environment. The robot's environment is a rectangular enclosed space, which is divided into a corridor and three small rooms. The rooms are equipped with bookshelves, tables, and ladders. The robot moves freely in the environment through small doors between the corridor and each room.

2. The ROS-based autonomous navigation robot according to claim 1, characterized in that: The SLAM of the mobile robot is composed of the classic SLAM method architecture, Rao-Blackwellised particle filter algorithm, gmapping algorithm, and gmapping algorithm mapping simulation. The classic SLAM method architecture uses a filter to perform state estimation on the collected environmental data and the measured robot motion model, and then obtains map data and robot posture. The Rao-Blackwellised particle filter algorithm regards the system state and measurement information as random variables, which conforms to the actual situation of mobile robot SLAM and is suitable for nonlinear, non-Gaussian mobile robot systems, avoiding any approximate linearization processing. The gmapping algorithm can simultaneously perform robot positioning and mapping by subscribing to sensor information and completing parameter configuration. The gmapping algorithm mapping simulation collects radar data of the unfamiliar environment around the robot to generate a grid map.

3. The ROS-based autonomous navigation robot according to claim 1, characterized in that: The mobile robot navigation algorithm research is based on the multi-sensor fusion algorithm of the extended Kalman filter and the ACML positioning algorithm and the mobile robot path planning algorithm; The adaptive Monte Carlo positioning algorithm in the multi-sensor fusion algorithm of the extended Kalman filter obtains the optimal final strategy through continuous strategy iteration. The characteristic of the MC control algorithm is to obtain the phased task according to the current strategy, and then update the value through all states in the multi-task, and then reach the convergence of the average value of the state value function through iteration. The AMCL positioning algorithm of ROS in the multi-sensor fusion algorithm of the extended Kalman filter is used for the global positioning of the robot in the map; The ACML positioning algorithm is a topic subscribed and published in ROS, and the AMCL algorithm simulation in the multi-sensor fusion algorithm of the extended Kalman filter is a robot positioning simulation and a posture tracking simulation; The mobile robot path planning algorithm is composed of the A* global path planning algorithm, the move_base navigation framework, and the robot path planning simulation; The mobile robot hardware platform is built to verify and analyze the robot's SLAM, path planning, and navigation algorithms, and uses the Raspberry Pi version 4B, in which the Ubuntu 18.04 operating system is installed and the ROS framework is encapsulated in the system, and the function package required for the experiment is installed as the top-level control platform of the robot, in which the robot's SLAM, positioning, and path planning are completed, and TM32F405RGT6 is used as the MCU of the underlying control board. The functions used include general I / O, general timer, general serial port, timer orthogonal decoding, timer PWM, IIC bus, and ADC voltage sampling; The ROS Mecanum wheel mobile robot platform is a robot, and the length, width and height of the robot are 30 cm, 24 cm and 18 cm respectively.

4. The ROS-based autonomous navigation robot according to claim 1, characterized in that: The robot SLAM experiment places the robot at the initial position, controls the robot to move in the room, and enables the laser radar to scan the entire room. After the scanning is completed, the generated map, that is, the robot can generate a complete room map. The room wall lines shown in the map are smooth, the internal environment details are clear, and the constructed map environment is highly consistent with the actual environment. The AMCL indoor positioning experiment uses the AMCL algorithm to position the robot. The robot positioning process, where the red arrow is the particle swarm generated by the algorithm, and the arrow direction indicates the posture of the robot; the red dot is the boundary position of the object displayed by the laser radar; The path planning and navigation experiment uses the move_base navigation algorithm to perform a path planning experiment on the robot, and experimentally analyzes the global path planning and local path planning of the robot in an actual scenario.

Images