Eco-Driving Longitudinal Control Strategy Based on Preceding Vehicle Acceleration Prediction North China University of Technology, Beijing Key Laboratory

Format:

Book

Conference/Event

Author/Creator:

Luo, Shije, author.

Contributor:

Zhao, Qi

Conference Name:

2025 International Conference on Intelligent Transportation and Future Mobility (ITFM2025) (2025-04-11 : Guilin, China)

Language:

English

Physical Description:

1 online resource cm

Place of Publication:

Warrendale, PA SAE International 2025

Summary:

Automatic driving technology can achieve precise control of the vehicle. Compared with manual driving, it can greatly avoid bad driving behaviors such as rapid acceleration, rapid deceleration, and idle driving, more stable, efficient and safer control of vehicles, thus reducing energy consumption and pollution emissions, has great potential for eco-driving. Previous research on eco-driving car-following strategy is usually based on the current vehicle state. However, the real driving scene is extremely complex and changeable, which makes the existing research easy to fall into the dilemma of local optimal solution when dealing with complex long-term planning tasks, and it is difficult to gain comprehensive insight into the path of global optimal solution. According to the literature, bad driving behaviors such as rapid acceleration and rapid deceleration have a great impact on the energy consumption and emissions of vehicles, in order to realize eco-driving, planning control method should be used to explore the range optimal strategy to approach the global optimal. Therefore, this paper discusses an eco-driving car-following strategy for autonomous vehicles based on the acceleration prediction of the leading vehicle, in this study, a Transformer-based acceleration prediction model for the leading vehicle is constructed, which uses the historical state time series data of the leading vehicle to predict its future state, so as to provide the future trajectory change trend of the leading vehicle for the following vehicle, the current state and the predicted state (through the feature fusion module) are used as decision variables for eco-driving, and the decision is adjusted according to the state change trend to achieve smooth driving, the strategy is suitable for the scenario where an autonomous vehicle follows a human-driven vehicle

Notes:

Vendor supplied data

Publisher Number:

2025-99-0445

Access Restriction:

Restricted for use by site license