Effective identification of drivers' driving styles ensures the safe operation of vehicles and enhances vehicle energy efficiency. Currently, research on the recognition of driving styles in battery electric vehicles is relatively limited, compounded by their unique characteristics in energy regeneration, resulting in significant differences in the internal data distribution compared to other powertrain systems. In this study, we introduce internal indicators such as torque and electricity to address the issue of relying solely on lagging external indicators, thus improving real-time recognition capabilities. To accommodate the distinctive data distribution patterns of battery electric vehicles, we employ the DBSCAN algorithm to address irregular cluster shapes. By leveraging sliding time windows and Long Short-Term Memory (LSTM) recurrent neural networks, we achieve real-time recognition of driving styles with an accuracy rate of 98.21%. In conclusion, this research bridges the gap in the field of driving style recognition for battery electric vehicles, providing robust support for practical applications.
A Driving Style Recognition Method for Battery Electric Vehicle Drivers
2023-10-28
1148814 byte
Conference paper
Electronic Resource
English
Electric vehicle braking energy recovery method based on driving style recognition
European Patent Office | 2024
|Hybrid electric vehicle control method and system based on driving style recognition
European Patent Office | 2020
|Driving style recognition model training and driving style recognition method, device and equipment
European Patent Office | 2024
|