Driving Condition Recognition Based on HMM for HEV Control

This paper presents a methodological approach for driving pattern recognition and prediction for Hybrid Electric Vehicle (HEV) control. In this approach, based on data collection in the real traffic conditions, several driving patterns are classified. These driving patterns represent different traffic conditions. Markov chain modeling has been used for traffic condition prediction. The driving pattern prediction can then be utilized for the HEV adaptive control. The simulation results are presented to investigate the effectiveness of the approach for the driving pattern prediction.

Keywords:

microtrip , traffic condition , driving pattern , Markov chain , recognition , prediction