Alberto Bertone
Abstract
This thesis deals with the study and simulation of methods to identify levels of drowsiness in drivers of motor vehicles during prolonged driving.
Fatigue and distraction are the cause of a large number of road accidents in the world. The causes of high fatigue during driving are mainly: lack of rest, excessive driving times, travel at night and so drowsiness. The vast majority of drivers do not recognize or underestimate the limitations of the ability to react due to fatigue. In fact, generally you continue the journey despite the arrival of clear symptoms of tiredness. Development of driving aids is of current interest in order to reduce critical situation. Specifically for the drowsiness, the aim is to warn the driver in time during travel, recognizing his degree of fatigue in advance.
In this document, in addition to an overall view of the existing techniques for identifying the drowsiness, it is presented a drowsiness model able to replicate different drowsy drivers according to their level of sleepiness. The main objective was the study and design of the driver model, simulating at the end the effective trajectory covered by the different car users in a standardized double lane change manoeuvre and in a longer open track.
Obiettivo Tesi
Realizzare un modello che permetta di simulare il comportamento di un autista con i 5 livelli di stanchezza definiti dalla letteratura.
Metodologia di ricerca
Stato dell’arte attraverso articoli e modelli Simulink sviluppati da altri enti di ricerca.
Conclusioni
Si è dimostrato però che è possibile definire un modello di guida da perfettamente vigile ad uno fortemente stanco.
Sviluppi futuri
Un primo modello di simulazione è stato realizzato, ma risulta necessario elimanare alcune instabilità numeriche.