Particle Filter Localization Simulation

Let us assume the following: You want a robot that will fetch something from somewhere or should take something to someplace. In order for this mobile robot to know if it reached the goal, or to plan a path to the goal, it needs to know where it is.

This 'knowing where it is' is known as localization. One method used for localization is Particle Filters (Monte Carlo Localization) and this simulation implements particle filter localization.

The simulation consists of RED particles, a GREEN robot and BLUE landmarks (add landmarks by LEFT clicking on the map). By moving the robot around using the w,a,d keys you will see how the particles cluster around the robot, essentially showing where the robot is. (NOTE: If you do not add any landmarks the robot will not be able to localize.)



Leave a Reply

Your email address will not be published. Required fields are marked *