Il full-text non è disponibile per scelta dell'autore.
(
Contatta l'autore)
Abstract
This thesis investigates the application of modern and advanced optimal control techniques within the realm of an innovative sports application. The application involves the use of an autonomous electric go-kart towing a plexiglass airshield, utilized to isolate Olympic athletes performing the 100m event from the drag influence during the overspeed training. The primary objective of this thesis is the deployment of a suitable controller for autonomously regulating the go-kart and the airshield in response to the position and the velocity of the runner. The main contribution of this work lies in the development of different controllers to address the desired task: a Gain Scheduling Linear Quadratic Regulator and a Model Predictive Controller. However, since a linear kinematic model has been used as the foundation for the controllers design, the real dynamic behaviour of the go-kart with the airshield is not perfectly described. The developed controllers are strongly model-based, and non-perfectly modeled dynamics effects or unknown disturbances could lead to suboptimal performance of the control architecture.
To address this issue, an Offset-free Model Predictive Control scheme has been introduced as a third control formulation. To demonstrate the effectiveness of the implemented controllers, they have been tested utilizing both Python simulations and hardware-in-the-loop tests. A comparison of the controllers performances is presented and analyzed within the framework of Python simulations utilizing data taken from 100m professional athletes' competitions. For the hardware-in-the-loop tests, an implementation based on the ROS 2 environment has been conducted. Several on-field experiments have been carried out in real-world settings to understand the performances of the controllers when operating with information coming from sensor measurements. This application holds significant potential for everyday use during training phases on the track and field racetrack.
Abstract
This thesis investigates the application of modern and advanced optimal control techniques within the realm of an innovative sports application. The application involves the use of an autonomous electric go-kart towing a plexiglass airshield, utilized to isolate Olympic athletes performing the 100m event from the drag influence during the overspeed training. The primary objective of this thesis is the deployment of a suitable controller for autonomously regulating the go-kart and the airshield in response to the position and the velocity of the runner. The main contribution of this work lies in the development of different controllers to address the desired task: a Gain Scheduling Linear Quadratic Regulator and a Model Predictive Controller. However, since a linear kinematic model has been used as the foundation for the controllers design, the real dynamic behaviour of the go-kart with the airshield is not perfectly described. The developed controllers are strongly model-based, and non-perfectly modeled dynamics effects or unknown disturbances could lead to suboptimal performance of the control architecture.
To address this issue, an Offset-free Model Predictive Control scheme has been introduced as a third control formulation. To demonstrate the effectiveness of the implemented controllers, they have been tested utilizing both Python simulations and hardware-in-the-loop tests. A comparison of the controllers performances is presented and analyzed within the framework of Python simulations utilizing data taken from 100m professional athletes' competitions. For the hardware-in-the-loop tests, an implementation based on the ROS 2 environment has been conducted. Several on-field experiments have been carried out in real-world settings to understand the performances of the controllers when operating with information coming from sensor measurements. This application holds significant potential for everyday use during training phases on the track and field racetrack.
Tipologia del documento
Tesi di laurea
(Laurea magistrale)
Autore della tesi
Cutini, Giulia
Relatore della tesi
Correlatore della tesi
Scuola
Corso di studio
Orientamento
PERCORSO STUDENTI CON CARENZA FORMATIVA
Ordinamento Cds
DM270
Parole chiave
Optimal Control,Model Predictive Control,Autonomous vehicle
Data di discussione della Tesi
18 Marzo 2024
URI
Altri metadati
Tipologia del documento
Tesi di laurea
(NON SPECIFICATO)
Autore della tesi
Cutini, Giulia
Relatore della tesi
Correlatore della tesi
Scuola
Corso di studio
Orientamento
PERCORSO STUDENTI CON CARENZA FORMATIVA
Ordinamento Cds
DM270
Parole chiave
Optimal Control,Model Predictive Control,Autonomous vehicle
Data di discussione della Tesi
18 Marzo 2024
URI
Gestione del documento: