/python_markov_localization

Python scripts to calculate a robots position using markov-localization

Primary LanguagePythonMIT LicenseMIT

Markov-Localization

This repository contains an implementation of the markov-localization-algorithm written in python.

To demonstrate its effectiveness, a robot is placed in a map with obstacles. The robot takes 8 steps and with each step, its new position is calculated. In between the steps, the robot is kidnapped and placed at a different location to show the robustnes of the algorithm in such cases.

Example

The robots steps are the following:

# Direction Measured distance to wall
1 Right 0
2 Up 0
3 Left 0
4 Down 0
5 Right 0
6 Top 1
7 Left 5
8 Down 1

After step 4, the robot is kidnapped from position (0,0) to position (6, 4).

Below, the resulting probability-grids as well as the current position of the robot is being shown.

Result