Hao Chen
Enhancing Road Safety and Energy Efficiency through Driving Behavior Detection Using Machine Learning Methods.
Rel. Angelo Bonfitto. Politecnico di Torino, Corso di laurea magistrale in Automotive Engineering (Ingegneria Dell'Autoveicolo), 2024
|
PDF (Tesi_di_laurea)
- Tesi
Licenza: Creative Commons Attribution Non-commercial No Derivatives. Download (22MB) | Preview |
Abstract: |
A significant rise in traffic fatalities linked to aggressive driving behaviors has been noted, accentuating the imperative need for research in this domain. Hence, the de- tection of aggressive driving is increasingly advocated as a strategy not only to alert drivers about their perilous behaviors but also to potentially diminish the incidence of accidents. In addition, driving style significantly affects the fuel efficiency of traditional internal combustion engine vehicles, and modifying driving behavior can effectively improve fuel economy. Compared to traditional vehicles, driving style has a greater impact on the range of electric vehicles. In parallel, the widespread adoption of driving simulators in the automotive sector has unveiled their profound advantages, especially in terms of repeatability in a controlled environment, helping researchers significantly reduce the time and cost of development. Furthermore, driving simulators also provide a safe platform for testing new technologies and evaluating driver behavior in various scenarios. This thesis presents a method for identifying aggressive driving by analyzing electric vehicle dynamics data (such as speed, acceleration, and steering angle) collected from simulation scenarios using the software SCANeR™Studio. The algorithm uses iterative density-based spatial clustering of noise applications (an unsupervised learning technique) to cluster aggressive driving maneuvers and sub- classify driving behaviors based on comfort, safety, and efficiency. In addition, these labeled data are used to train Bayesian optimization-based long short-term memory neural network and a random forest model. The research results demonstrate that excels in accurately identifying energy- efficient driving behaviors and aggressive driving behavior, with a F-score 0.992 and 0.869, showing great potential in enhancing road safety as well as promoting vehicle energy conservation and sustainable driving practices. |
---|---|
Relators: | Angelo Bonfitto |
Academic year: | 2023/24 |
Publication type: | Electronic |
Number of Pages: | 75 |
Subjects: | |
Corso di laurea: | Corso di laurea magistrale in Automotive Engineering (Ingegneria Dell'Autoveicolo) |
Classe di laurea: | New organization > Master science > LM-33 - MECHANICAL ENGINEERING |
Aziende collaboratrici: | Politecnico di Torino |
URI: | http://webthesis.biblio.polito.it/id/eprint/31582 |
Modify record (reserved for operators) |