December 31, 2019

Path planning in crowded places using reinforcement learning

Panayiota Alexoudi
  • November 1, 2019
  • December 31, 2019
  • Internship
  • Aristotle University of Thessaloniki, Faculty of Sciences, School of Mathematics
  • Panayiota Alexoudi
  • Stasinos Konstantopoulos
  • The AI Lab

Abstract:

The internship is about the movement of a robotic platform in presence of multiple moving obstacles. The data come from a stationary platform that records the distance traces of people in motion. They will be implemented:

Convert from the original data to tables in Cartesian coordinates.

A method that calculates whether a collision with an obstacle has occurred given a position of the platform in the same axle system.

A method that calculates the position of the moving platform in the same axis system, assuming a platform motion algorithm.

Reinforcement learning experiment that improves the initial algorithm of movement that moves straight to the target position without taking into account the obstacles.

Skip to content