Modelling, control and simulation of a 4-legged robot for obstacle avoidance

Abstract

This project aims to develop a simulation of a quadruped robot for obstacle avoidance. The goal of the simulation is having the robot to avoid obstacle while performing SLAM using 2D LiDAR sensor. Additionally, if the robot encounters obstacle that is not detected by the LiDAR sensor, the robot is capable to respond to the situation. The simulation is built upon ROS2, Gazebo, Rviz, and Nav2. The environment is built in Gazebo, and the illustration of the mapping is shown in Rviz. The robot is equipped with a 2D-LiDAR sensor to provide the robot with the capability of obstacle avoidance. IMU sensor is implemented to observe the orientation of the robot, and check if the robot body is tilted in the case when the robot is climbing slope or obstacle. The overall simulation has successfully implemented and achieved the following: The integration of Gazebo and Rviz for complete simulation testing environment, using 2D LiDAR to scan and constructing a map, using Nav2 plugin to perform recovery action when facing small obstacle, and using IMU sensor to implement a close loop feedback controller.