Prediction of Electric Vehicles Schedulable Capacity Based on Graph Convolution Network

Conference: PCIM Asia 2024 - International Exhibition and Conference for Power Electronics, Intelligent Motion, Renewable Energy and Energy Management
08/28/2024 - 08/30/2024 at Shenzhen, China

doi:10.30420/566414067

Proceedings: PCIM Asia 2024

Pages: 5Language: englishTyp: PDF

Authors:
Mao, Meiqin; Wu, Jixun; Yang, Cheng; Wang, Yuanyue; Zhu, Minglei; Hatziargyriou, Nikos

Abstract:
Spatial-temporal prediction of the electric vehicles schedulable capacity is necessary for their integration into power grid management. This paper proposes a spatial-temporal prediction model based on graph convolution. Each charging station within a region is considered a node, the connection relationship is determined based on the spatial distance between charging stations, obtaining the charging station connectivity graph. Subsequently, using the graph convolution network to explore spatial features from the charging data, and the gate recurrent unit is used to mine temporal features, resulting in spatial-temporal predictions of the schedulable capacity of electric vehicles. Finally, the proposed method is validated and compared with other prediction methods.