This paper presents a classification model to identify abnormal driving behavior on roads. Vehicle dynamics is considered. A LSTM recurrent neural network model-based is applied. The vehicle dynamics features are measured by smartphone inertial sensors. The real data obtained from the GPS, accelerometer, and gyroscope are used to classify the driving maneuvers. A conventional two-lane and a highway roads located in the Madrid Region, Spain, are used for this research. The results obtained with the proposed model are promising and suggest that this intelligent system can be used to warn drivers of a defective or distracted maneuver in real time, aiming at a safer and more comfortable driving.
Abnormal Driving Behavior Identification Based on Naturalistic Driving Data Using LSTM Recurrent Neural Networks
Lect. Notes in Networks, Syst.
International Workshop on Soft Computing Models in Industrial and Environmental Applications ; 2022 ; Bilbao, Spain September 05, 2022 - September 07, 2022
17th International Conference on Soft Computing Models in Industrial and Environmental Applications (SOCO 2022) ; Kapitel : 42 ; 435-443
2022-10-12
9 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Driving Style Clustering using Naturalistic Driving Data
Transportation Research Record | 2019
|Studying Driving Behavior on Horizontal Curves using Naturalistic Driving Study Data
Transportation Research Record | 2018
|