With the continuous advancement of technology and growing global awareness of environmental issues, the demand for electric vehicles (EVs) and other new energy vehicles (NEVs) has been increasing at an unprecedented rate. This surge has heightened the requirements for optimized charging load. However, a notable limitation in current research lies in its predominant focus on broad, macro-level traffic flow models. These models often struggle to accurately represent fine-grained simulations of individual vehicles and frequently overlook the application of vehicle-to-grid (V2G) technology. These limitations hamper the safe operation of the power grid and the efficient utilization of resources. Therefore, by incorporating the application of V2G and employing refined simulation models, this research aims to replicate the charging patterns of electric vehicles within a realistic grid setting, thereby delving into the V2G load simulation. Furthermore, this paper proposes the utilization of deep learning model to fit and predict the V2G loads from data generated by refined traffic flow simulations. Integrating deep learning techniques is expected to significantly reduce simulation time while improving the performance and usability of the simulation results. This approach is designed to provide reliable technical support and assurance for the widespread adoption of electric vehicles in the future.
Deep Learning-based Vehicle-to-Grid Loads Simulation and Prediction Considering Microscopic Traffic Behaviors
2024-10-10
4683534 byte
Conference paper
Electronic Resource
English
Evaluation of Driving Behaviors in Traffic Flow using Microscopic Traffic Simulation
British Library Conference Proceedings | 2009
|Microscopic Traffic Simulation by Cooperative Multi-agent Deep Reinforcement Learning
ArXiv | 2019
|Vehicle trajectory reconstruction for freeway traffic considering lane changing behaviors
Taylor & Francis Verlag | 2025
|