Full text
| |
| Author(s): |
Rafael de Angelis Cordeiro
Total Authors: 1
|
| Document type: | Doctoral Thesis |
| Press: | Campinas, SP. |
| Institution: | Universidade Estadual de Campinas (UNICAMP). Faculdade de Engenharia Elétrica e de Computação |
| Defense date: | 2017-09-28 |
| Examining board members: |
Paulo Augusto Valente Ferreira;
Jose Reginaldo Hughes Carvalho;
Guilherme Augusto Silva Pereira;
Rafael Santos Mendes;
Alim Pedro de Castro Gonçalves
|
| Advisor: | Alessandro Correa Victorino; Ely Carneiro de Paiva; Paulo Augusto Valente Ferreira; José Raul Carreira Azinheira |
| Abstract | |
Intelligent ground vehicles have received special attention from the robotics community in the past years. Their applications are vast, including: new safety technologies adopted by commercial passengers cars worldwide; harvest autonomous machines; interplanetary exploration robots, among others. Despite the variety of applications, their principle of movement is the same: the interaction between the vehicle and the ground. In this thesis, we are interested in the interactions between the vehicle, their tires, and the ground. More precisely, we are interested in how to measure interaction forces and describe their influences on the vehicle behavior, especially in extreme conditions such as the lack of adherence, irregular terrains, excessive speed, intense cornering, etc. To answer these questions, we address the problem using an applied control point of view. Firstly, the system to be controlled - the vehicle - is modeled using differential equations. Some hypotheses are introduced, and an analytical approach of its dynamics is used to create several vehicle models with different levels of complexity. Control strategies need to accurately sense the system dynamics. Some variables are difficult to measure or need expensive sensors. The most difficult variables to be measured are exactly the forces acting on the tire-ground interface. Here, we present a new delayed interconnected cascade-observer structure to accurately estimate the forces acting on each tire. The results obtained are compared with real data acquired from an actual vehicle platform. As control goal, the vehicle must execute maneuvers with precision. To determine the ways the vehicle can behave, standard nonlinear system analyses of vehicle state-space models are developed. Such analyses consider different characteristics of the vehicle: the traction forces differential, the combined tire forces, the vertical forces distribution, and the adherence condition. The analyses provide cornering conditions, including normal and drifting cornering, when the vehicle is submitted to different configurations of speed and steering. These conditions are used to synthesize controllers in a simulated environment. Results show that the controllers are capable of stabilizing the vehicle in cornering, even in drifting conditions (AU) | |
| FAPESP's process: | 14/06610-8 - Path tracking control of a four-wheel robotic vehicle under adverse terrain conditions |
| Grantee: | Rafael de Angelis Cordeiro |
| Support Opportunities: | Scholarships in Brazil - Doctorate |