MPC control of an autonomous wheelchair considering ride comfort

This thesis work concerns vehicle robotics, the main objective being the development of an algorithm for trajectory planning for autonomous vehicles. To ful ll these requirements, the technique utilized is the so-called Model Predictive Control, which allows to perform an optimal trajectory generati...

Descripción completa

Detalles Bibliográficos
Autor: Sánchez Moreno, Adrià
Tipo de recurso: tesis de maestría
Fecha de publicación:2018
País:España
Institución:Universitat Politècnica de Catalunya (UPC)
Repositorio:UPCommons. Portal del coneixement obert de la UPC
Idioma:inglés
OAI Identifier:oai:upcommons.upc.edu:2117/169071
Acceso en línea:https://hdl.handle.net/2117/169071
Access Level:acceso abierto
Palabra clave:Robotics
Robòtica
Àrees temàtiques de la UPC::Enginyeria mecànica::Fabricació::Disseny i accionaments de robots
Descripción
Sumario:This thesis work concerns vehicle robotics, the main objective being the development of an algorithm for trajectory planning for autonomous vehicles. To ful ll these requirements, the technique utilized is the so-called Model Predictive Control, which allows to perform an optimal trajectory generation while guaranteeing other control requisites such as obstacle avoidance, comfort or safety. The focus is to assure ride comfort for the passenger in an scenario of obstacle avoidance, based on previous work on the system. In this thesis, the vehicle is an electric wheelchair that receives reference values of velocity from a controller, which applying the MPC technique minimizes a suitable cost function based on the system and the geometry of the environment for a nite prediction horizon. The requirements of ride comfort while ful lling obstacle avoidance are based on previous study of human body vibrations in the literature, and their e ect to the comfort sensation of the passenger, emphasizing in determining this sensation through an objective method. The performance of the resulting control algorithm is evaluated by means of simulations to prove the e ectiveness of this approach to the problem.