Open AccessResearch on Electric Vehicle Charging Load Prediction Methods Combining Signal Noise Reduction and Time Series Modeling
Author(s) -
Liyun Liu,
Xiaomei Xu,
Jinsong Zhang,
Dong Li
Publication year - 2025
Publication title -
ieee access
Language(s) - English
Resource type - Magazines
SCImago Journal Rank - 0.587
H-Index - 127
eISSN - 2169-3536
DOI - 10.1109/access.2025.3574409
Subject(s) - aerospace , bioengineering , communication, networking and broadcast technologies , components, circuits, devices and systems , computing and processing , engineered materials, dielectrics and plasmas , engineering profession , fields, waves and electromagnetics , general topics for engineers , geoscience , nuclear engineering , photonics and electrooptics , power, energy and industry applications , robotics and control systems , signal processing and analysis , transportation
Accurate prediction of electric vehicle (EV) charging demand is crucial for grid stability, energy allocation, and charging infrastructure planning. This study introduces a hybrid deep learning model combining Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN), Convolutional Neural Networks (CNN), Bi-directional Gated Recurrent Units (BiGRU), and Attention Mechanism (AM) to address the volatility in charging load patterns. The proposed method begins with the application of CEEMDAN to decompose the data into intrinsic mode functions (IMFs) and residuals. Subsequently, it employs convolutional and pooling layers within the CNN to reduce the dimensionality of the data, thereby optimizing the input dimensions. The BiGRU is then utilized to analyze both long- and short-term temporal dependencies inherent in the load data features, facilitating enhanced feature extraction. Finally, the AM is introduced to prioritize significant components within the load data features, thereby improving the model's forecasting accuracy. To validate the prediction accuracy of proposed model, a comparative experiment is conducted comparing with five cutting-edge prediction models. The results demonstrate that proposed model achieves 11.37% average improvement in prediction accuracy. These findings confirm the reliability and effectiveness of the proposed model for user charging demand load prediction, showcasing its strong generalization ability and robustness.