logo SBA

ETD

Archivio digitale delle tesi discusse presso l’Università di Pisa

Tesi etd-10112023-134848


Tipo di tesi
Tesi di laurea magistrale
Autore
MANTOVA, LUCIA
URN
etd-10112023-134848
Titolo
Model based analysis of yacht's hotellerie consumption and optimization for sustainable user comfort by AI
Dipartimento
INGEGNERIA DELL'INFORMAZIONE
Corso di studi
INGEGNERIA ROBOTICA E DELL'AUTOMAZIONE
Relatori
relatore Prof. Saponara, Sergio
relatore Ing. Dini, Pierpaolo
Parole chiave
  • yacht
  • hotellerie
  • consumption
  • neural network
  • ffnn
  • HVAC
  • smart office
  • MATLAB
  • Simulink
  • AI
Data inizio appello
23/11/2023
Consultabilità
Non consultabile
Data di rilascio
23/11/2093
Riassunto
This thesis covers a relatively new topic, as no similar project has been found in the literature. In fact, a smart yacht model has been created from scratch using MATLAB and Simulink, primarily based on everything related to the smart office and smart buildings concept, with the necessary exceptions due to the nature of the vessel. Energy efficiency has been of paramount importance for this topic: it was necessary to consider every system (from the HVAC system installed in a cabin to individual appliances) capable of contributing to the energy flow of the entire yacht, taking into account energy input from the engine and solar panels as well. External conditions were also analyzed, along with how they could influence the energy required to power the yacht and satisfy the occupants on board. For example, how external temperature would impact the use of the air conditioning system or how weather conditions would affect the solar panels.
Additionally, at the request of the client Rossinavi, four different navigation modes were considered, from which users can choose (Regular, EcoGuest, EcoCross and Hibernation), along with various levels of privileges assignable to the users on the ship. Furthermore, the entire model has been constructed to incorporate an artificial intelligence component, Blue AI, capable of providing forecasts, supporting users, and intervening autonomously as needed. In particular, the predicted power consumptions are calculated using FFNNs, with an assigned time frame, and based on one or more user profile trends. All of this was made possible through collaboration with the company Videoworks and Rossinavi, which provided the luxury yacht SeaCat 40, on which the entire model was built.
File