This paper aims to investigate rider braking behaviors using a dataset of braking maneuvers derived from naturalistic riding data. Each braking event was fully characterized with experimental data. A set of descriptive parameters was defined to capture relevant information of the braking event and to facilitate the clustering process of braking behaviors. Naturalistic data of 5 riders were automatically processed to identify and characterize the braking events based on the given set of parameters. A preliminary descriptive analysis was performed to verify the presence of macro behaviors of riders. Subsequently, a Principal Component Analysis was performed to reduce problem dimensionality and support the cluster analysis on the dataset of a rider. The results indicated that a macro classification of riders is possible also based on a descriptive analysis. Nonetheless a cluster analysis sharply identified different behaviors of the rider, and thus provided a more solid basis for comparison of behavior among riders. In addition, the clusters revealed quantitative data that will be useful for the development of assistive systems.
http://hdl.handle.net/2158/1066353
Titolo: | Rider Behavioral Patterns in Braking Manoeuvres |
Autori di Ateneo: | |
Autori: | BALDANZINI, NICCOLO'; HUERTAS LEYVA, PEDRO; SAVINO, GIOVANNI; PIERINI, MARCO |
Anno di registrazione: | 2016 |
Rivista: | TRANSPORTATION RESEARCH PROCEDIA |
Titolo del libro: | Transportation Research Procedia |
Titolo del congresso: | Transport Research Arena 2016 Conference |
Luogo del congresso: | Varsavia |
Data del congresso: | 18-21 aprile 2016 |
Abstract: | This paper aims to investigate rider braking behaviors using a dataset of braking maneuvers derived from naturalistic riding data. Each braking event was fully characterized with experimental data. A set of descriptive parameters was defined to capture relevant information of the braking event and to facilitate the clustering process of braking behaviors. Naturalistic data of 5 riders were automatically processed to identify and characterize the braking events based on the given set of parameters. A preliminary descriptive analysis was performed to verify the presence of macro behaviors of riders. Subsequently, a Principal Component Analysis was performed to reduce problem dimensionality and support the cluster analysis on the dataset of a rider. The results indicated that a macro classification of riders is possible also based on a descriptive analysis. Nonetheless a cluster analysis sharply identified different behaviors of the rider, and thus provided a more solid basis for comparison of behavior among riders. In addition, the clusters revealed quantitative data that will be useful for the development of assistive systems. |
Handle: | http://hdl.handle.net/2158/1066353 |
Appare nelle tipologie: | 4a - Articolo in atti di congresso |
File in questo prodotto:
File | Descrizione | Tipologia | Licenza | |
---|---|---|---|---|
1-s2.0-S2352146516303659-main.pdf | Articolo definitivo | PDF editoriale | ![]() | Accesso pubblicoVisualizza/Apri |