| Title |
Deep Learning-based System Warning of Short Remaining Operational Times in Electric Vehicles |
| Authors |
(Sanghyeon Lee) ; (Taeo Kim) ; (Duckki Lee) ; (Sung Wook Park) |
| DOI |
https://doi.org/10.5573/IEIESPC.2020.9.10.345 |
| Keywords |
Warning; Operational time; Battery; Aging; Deep learning |
| Abstract |
To predict the remaining operational time of a battery is important in electric vehicle (EV) operation. However, the nonlinear characteristics of battery cells and the different driving patterns of people hinder that prediction. Furthermore, the aging characteristics of battery cells make predictions even more difficult. This paper presents a deep learning system that warns drivers that an EV may not be drivable in a short time. The training dataset reflects the nonlinearity of battery cells, random driving patterns, and the aging characteristics of battery cells, with values normalized to cover various forms of battery packs, which are combinations of battery cells. The performance of the proposed warning system shows around 99% accuracy for constant-speed driving situations and around 78% accuracy for random driving patterns, with the warning designed to be given three minutes before full battery discharge. |