Navigation d'un robot mobile en environnement inconnu utilisant les champs de potentiels artificiels

This project deals with the navigation of a mobile robot in an unknown environment. The aim is to develop a complete method that allows controlling a robot to reach its goal while avoiding unknown obstacles on its way using the least amount of computational or sensing resources. It is also aimed at minimizing the tuning period before the execution of the task at hand, by developing an objective approach that computes the required parameters in an automatic fashion.;A new form of repelling potential is proposed in order to reduce oscillations and to avoid conflicts when the target is close to obstacles. An additional force is proposed as well, allowing for a smoother trajectory around the obstacles.;With the new form of potentials, an objective computation method for the algorithm's parameters is proposed. It uses optimization criteria to determine the adequate value of those parameters.;The theory developed is tested by simulations and experimentation on a real robot. Simulations are done using Matlab in a controlled environment, while practical experimentation is done on a real robot (Pioneer 3AT) in a less then ideal real environment. The robot's model is also detailed in order to put the conditions of the experimentation into context.;The approach developed is based on the Artificial Potential Field (APF) method in which the target creates a virtual force that attracts the robot while obstacles create a virtual force that repels the robot. APF-based methods are interesting because they are simple to implement; however, they have some inherent limitations that create trajectory oscillations, among other things.