Implementation of Cluster Based Exploration Algorithm For Mobile Robot Using ROS and Gazebo

Abstract

Online path planning and obstacle avoidance algorithms help in rescue operations of natural and human made disasters. It is done by calculating the trajectory online which allows the robot to react to changes in environment and the obstacles it encounter. In this work, a novel algorithm for obstacle avoidance and exploration is developed and implemented using Robotic Operating System (ROS) and Gazebo. The code is run on MATLAB and the simulation result is obtained from Gazebo. ROS acts as an interface between Gazebo and MATLAB. VMware provides a common platform which links ROS and Gazebo. Cluster based exploration algorithm divides the recent sensor readings into four sets of data and decides the appropriate direction and velocity depending on the presence of obstacle.