| dc.contributor.advisor | Rus Pezzi, Joaquín | |
| dc.contributor.author | Bravo Moreno, Francisco | |
| dc.date.accessioned | 2024-07-30T10:51:07Z | |
| dc.date.available | 2024-07-30T10:51:07Z | |
| dc.date.issued | 2023 | |
| dc.identifier.citation | Bravo Moreno, F. (2023) Análisis del impacto técnico y económico de la aplicación de algoritmos de control basados en RL en nuevos usos de drones [Trabajo Fin de Máster, Universidad Loyola Andalucía] | es |
| dc.identifier.uri | https://hdl.handle.net/20.500.12412/6111 | |
| dc.description.abstract | There are many new devices that the advancement of technology has brought into our
daily lives. Concepts like smartphones, 5G, or drones were unknown to the general
population. However, we can say that the revolution that brought these terms to the
general public is already seen with some distance as we are being swept away by another
technological paradigm shift, artificial intelligence.
Artificial intelligence, the protagonist of countless newspaper headlines, is changing the
way we use devices. It is changing the relationship that society has with machines. It is a
new technological capability that has allowed for a new range of possibilities regarding
what we believe machines can and can´t do.
The objective of this study is to explore the possibility of implementing one of the
algorithms that embodies this technology in controlling the various actuators of a
quadrotor drone. It will analyse the impact it has on the operational philosophy of these
systems, the challenges and problems that may arise along the way. Also it will explore
the impact of integrating artificial intelligence into the drone market and the services
offered with them.
To achieve this, an overview of the control technology currently used in drone flight will
be provided, analysing its structure and performance. This will be followed by an
introduction to Reinforcement Learning, a branch of Machine Learning and it’s case for
drone control.
This performance improvement in the drone controlled by Reinforcement Learning will
be estimated through a Python simulation, comparing the speed of achieving different
objectives between a PID-controlled drone and one controlled by one of the described
intelligent algorithms.
The execution of these algorithms will require modifications to the hardware of the drone
control systems. These requirements will be presented and analysed in other to seek for
solutions.
The final part of the study will introduce the current reality of the industry. It will cover
the size of the drone market, its main fields of application, and the trends identified for
the upcoming years, along with the key players in the industry. | es |
| dc.language.iso | spa | es |
| dc.rights | Attribution-NonCommercial-NoDerivatives 4.0 Internacional | * |
| dc.rights.uri | http://creativecommons.org/licenses/by-nc-nd/4.0/ | * |
| dc.title | Análisis del impacto técnico y económico de la aplicación de algoritmos de control basados en RL en nuevos usos de drones | es |
| dc.type | masterThesis | es |
| dc.description.master | Máster Universitario en Ingeniería Industrial | es |
| dc.rights.accessRights | openAccess | es |