Home / Current Issue / Paper 1718001
A Systematic Review and Proposed Hybrid SSA-CEEMDAN-Attention Framework for Stochastic Electric Vehicle Charging Demand Forecasting
Subject area: Science,Engineering and Technology · Area of research: Computer Science
DOI: https://doi.org/10.64388/IREV9I11-1718001
Abstract
The exponential surge in Electric Vehicle (EV) adoption presents a high-dimensional, spatio-temporal stochastic optimization problem that threatens the operational stability of contemporary power distribution networks. Precise short-term demand forecasting is a mechanical necessity for integrating charging stations (EVCS) with volatile renewable energy sources. Existing literature frequently fails to synchronize the “Double Uncertainty” inherent in coupled behavioral and meteorological patterns, while simultaneously exceeding the computational thresholds of decentralized edge devices. This study synthesizes 25 recent works published between 2019 and 2026 and proposes a novel hybrid SSA-CEEMDAN-Attention-BiLSTM architecture. Preliminary benchmarking indicates that the proposed selfoptimizing denoising framework achieves a 15% reduction in Mean Absolute Percentage Error (MAPE) under high-volatility conditions, providing a computationally efficient blueprint for the 2030 smart grid transition.
Keywords
Electric Vehicle (EV), Load Forecasting, CEEMDAN, Attention Mechanism, Smart Grid, Hybrid Models, Deep Learning, Optimization
References
[1] V. Dhanawat et al., “Electric Vehicles Charging Station Load Forecasting Integration With Renewable Energy Sources,” IEEE Open Journal of Vehicular Technology, vol. 6, pp. 1–15, 2025.
[2] T. Rasheed et al., “Improving the Efficiency of Deep Learning Models Using Supervised Approach for Accurate Electric Vehicle Charging Load Forecasting,” IEEE Access, vol. 11, pp. 102261–102275, 2023.
[3] E. Fahnbulleh et al., “An AI Driven, Reliability-Based Planning Framework for Electric Vehicle Charging Infrastructure,” IEEE Access, vol. 13, pp. 12890–12905, 2025.
[4] L. Liu et al., “Research on EV Charging Load Prediction Methods Combining Signal Noise Reduction and Deep Learning,” IEEE Access, vol. 13, pp. 8840–8855, 2025.
[5] D. Ny and C. Jeenanunta, “Optimizing Power System Reliability and Carbon Emissions With a Fuzzy Unit Commitment,” IEEE Access, vol. 12, pp. 107421–107435, 2024.
[6] L. Liu et al., “Signal Noise Reduction Techniques for High-Precision EV Load Prediction,” IEEE Access, vol. 13, pp. 8840–8855, 2025.
[7] G. Gruosso et al., “Uncertainty-Aware Computational Tools for Power Distribution Networks,” IEEE Access, vol. 7, pp. 28916–28930, 2019.
[8] W. Ning and G. Li, “New Energy Electric Vehicle Charging Load Forecasting Based on the SSA-CNN-LSTM Model,” Journal of Electrotechnology, Electrical Engineering and Management, vol. 8, no. 1, pp. 96–103, 2025.
How to cite this paper
@article{1718001,
author = {Ravi Kishan Varma, Dr. S. Balamurugan},
title = {A Systematic Review and Proposed Hybrid SSA-CEEMDAN-Attention Framework for Stochastic Electric Vehicle Charging Demand Forecasting},
journal = {Iconic Research And Engineering Journals},
year = {2026},
volume = {9},
number = {11},
pages = {2959-2962},
issn = {2456-8880},
url = {https://www.irejournals.com/formatedpaper/1718001.pdf},
abstract = {The exponential surge in Electric Vehicle (EV) adoption presents a high-dimensional, spatio-temporal stochastic optimization problem that threatens the operational stability of contemporary power distribution networks. Precise short-term demand forecasting is a mechanical necessity for integrating charging stations (EVCS) with volatile renewable energy sources. Existing literature frequently fails to synchronize the “Double Uncertainty” inherent in coupled behavioral and meteorological patterns, while simultaneously exceeding the computational thresholds of decentralized edge devices. This study synthesizes 25 recent works published between 2019 and 2026 and proposes a novel hybrid SSA-CEEMDAN-Attention-BiLSTM architecture. Preliminary benchmarking indicates that the proposed selfoptimizing denoising framework achieves a 15% reduction in Mean Absolute Percentage Error (MAPE) under high-volatility conditions, providing a computationally efficient blueprint for the 2030 smart grid transition.},
keywords = {Electric Vehicle (EV), Load Forecasting, CEEMDAN, Attention Mechanism, Smart Grid, Hybrid Models, Deep Learning, Optimization},
month = {May},
doi = {https://doi.org/10.64388/IREV9I11-1718001}
}