Development of a Dynamical Model for Retina-based Control of an RC Monster Truck
Author(s): Robin Ritz |
Conference/Journal: Semester/Bachelor Thesis, FS 08 |
Abstract: At the Institute of Neurinformatics (INI), a shared institute of the University of Zurich (UZH) and the Swiss Federal Institute of Technology Zurich (ETHZ), a dynamic vision sensor for fast visual processing has been developed. The dynamic vision sensor uses a silicon retina to detect contrast changes in a scene. An ongoing project is to develop a controller based on a silicon retina for a small robotic monster truck. The goal of the project is to drive at high speed along a drawn route. In this semester project, a dynamical model of the RC monster truck is derived and a simulation environment to test retina-based control strategies is developed. The simulation is based on Blender, a powerful open-source modeling environment. Further, first control strategies for event-based lateral control are implemented and tested in the simulation, as well as on the real vehicle. At the end of this semester project, the truck is able to follow a chalk drawn line at a little more than walking pace. Supervisors: Prof. M. Morari, Dr. T. Besselmann | Year: 2008 |
Type of Publication: (13)Semester/Bachelor Thesis | |
Supervisor: M. Morari | |
% Autogenerated BibTeX entry @PhdThesis { Xxx:2008:IFA_3190 } |