From AIRWiki

Jump to: navigation, search
AGW - Automatic Guided Wheelchair
Image of the project AGW
Short Description: Control system design of an electric wheelchair for autonomous drive with obstacle avoidance
Tutor: MatteoMatteucci (, Marcello Farina (), Luca Bascetta ()
Students: MauroGabellone (, EugenioCeravolo (
Research Area: Robotics
Research Topic: Robot development
Start: 2015/11/??warning.pngThe date "2015/11/??" was not understood.
Status: Active
Level: Ms
Type: Thesis

The goal of this master thesis project is to take over the control of a wheelchair (designed to be controlled through a lever) in order to make it an autonomous vehicle. The first stage will concern the identification of the system, which will provide the relation between the control input (lever position) and wheels actuators. The project will be carried out step by step: as a first stage, the control will merely involve trajectory generation and following (by means of Model Predictive Control algorithms). Then, obstacles will be placed in the environment and the control problem will focus on obstacle avoidance (either fixed or moving). Two kinds of sensors will be available: a camera sensor (which will provide almost deterministic measurements) will be used in the first stages of the project, while two laser sensors (and their stochastic measurements) embedded on the wheelchair will be used in the latter stages as a more realistic scenario. The purpose is to design a robust control system which would take care of the passenger comfort and energetic optimization given the task to be carried out.