Project

Overview

The objective of this exercise was to implement, analyze, and understand the working of a version of the Particle Filter Localization Algorithm (called Monte Carlo Localization, MCL). The project was implemented through a simulation, specifically using the ROS-Stage and Rviz platforms and it was written in Python programming language. In order to write the localisation node, we used the skeleton code provided https://github.com/JonFreer/pf_localisation. Then, we used the map provided for the practice exercise in the "socspioneer" package. In addition to the steps of the Particle Filtering (i.e., motion update, sensor update, and re-sampling), we also investigated methods for obtaining a single estimate of the robot’s pose at each time step and we dealt with the ”kidnapped robot” problem.