Politecnico di Torino (logo)

Vehicle path prediction for safety enhancement of autonomous driving

Andrea Mancini

Vehicle path prediction for safety enhancement of autonomous driving.

Rel. Stefano Alberto Malan, Daniele Brevi, Francesca Pacella. Politecnico di Torino, Corso di laurea magistrale in Mechatronic Engineering (Ingegneria Meccatronica), 2021

PDF (Tesi_di_laurea) - Tesi
Licenza: Creative Commons Attribution Non-commercial No Derivatives.

Download (8MB) | Preview

In the last few years the concept of autonomous vehicle has developed significantly and has become one of the major topics of the automotive field. Among all the functions that a self - driving vehicle must have, the ability to correctly interpret the sensors information and exploit them for trajectory prediction is fundamental. This thesis work collocates itself exactly in this framework, and it is developed around two main topics. First of all, a research on motion models is conducted in order to establish which one can better describe the analysed scenario. Then, the chosen model, called CTRA model, is simulated with real sensor data with the purpose of obtaining a trajectory of the same shape of the authentic one. The second phase of the project has the goal to simulate a real - time scenario and consists in the combination of the CTRA model and the Unscented Kalman Filter for trajectory prediction purposes. Two different case of application are examined and compared: the first analysis is conducted from the vehicle's point of view, while, in the second case, the trajectory prediction is obtained thanks to the data acquired from a camera located on the side of the road. The achieved results have shown that a more accurate prediction is obtained when the correction done by the filter is based on a greater number of measured variables, but they have also revealed that the assumptions proposed by the utilized model are more suitable for short - term prediction, while they become a burden for longer prediction horizons.

Relators: Stefano Alberto Malan, Daniele Brevi, Francesca Pacella
Academic year: 2020/21
Publication type: Electronic
Number of Pages: 87
Corso di laurea: Corso di laurea magistrale in Mechatronic Engineering (Ingegneria Meccatronica)
Classe di laurea: New organization > Master science > LM-25 - AUTOMATION ENGINEERING
Aziende collaboratrici: FONDAZIONE LINKS
URI: http://webthesis.biblio.polito.it/id/eprint/18264
Modify record (reserved for operators) Modify record (reserved for operators)