Real‐time driving style classification based on short‐term observations

Vehicle behaviour prediction provides important information for decision-making in modern intelligent transportation systems. People with different driving styles have considerably behaviours and hence exhibit tendency. However, most existing methods do not consider the tendencies apply same model to all vehicles. Furthermore, of driver classification rely on offline learning that requires a long observation history are suitable real-time analysis. To facilitate personalised models can potentially improve vehicle prediction, authors propose an algorithm classifies drivers into styles. The only data from short window it is more applicable online applications compared require term observation. Experiment results demonstrate proposed achieve consistent provide intuitive interpretation statistical characteristics styles, which be further used prediction.