← Back to all news
arXiv Machine Learning By Graeme Kelly, Emilio J. Palacios-Garcia, Barry P. Hayes

Synthetic Electric Vehicle Charging Session Generation Using a Conditional Variational Autoencoder

Read the original on arXiv Machine Learning →

The paper introduces a conditional variational autoencoder (CVAE) to generate synthetic electric vehicle (EV) charging sessions from real transaction-level data. It trains on engineered features such as plug‑in duration, charging duration, delivered energy, charging delay, and cyclical time‑of‑week, conditioning on day of week and managed charging status. Evaluation shows the synthetic data preserves key statistical properties and supports predictive modelling tasks via a Train‑on‑Synthetic‑Test‑on‑Real protocol.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Machine Learning.

Related stories

arXiv Machine Learning
Aug 31

Conditional Diffusion Models for Energy-Efficient Driving

The paper presents a conditional diffusion model that generates electric vehicle battery‑current profiles conditioned on route features such as velocity and ambient temperature. Using a latent conditioning encoder and a temporal 1D U‑Net denoising backbone, the model produces realistic current trajectories that capture both the overall envelope and sharp transient events. On a dataset of 12,000 trips from nine vehicles, the model achieves a Wasserstein distance of 0.0029, outperforming direct condition injection by 89.1% in Wasserstein distance and 52.8% in MAE.

By Hemanth Neelgund Ramesh, Andr\'e Snoeck, Chyi-Fu Hong, Shijing Sun
diffusion
More like this →
arXiv AI
Aug 26

A Behavior-Guided Online Probabilistic Forecasting Method for Electric vehicle Charging Loads

The paper introduces a behavior-guided online probabilistic forecasting framework for electric vehicle charging loads that captures both persistent station-specific patterns and recent behavioral changes through a dual-timescale representation. It employs semantic encoding of behavioral shifts to adapt forecasts in a drift-aware manner and uses a delayed-feedback mechanism to maintain temporal consistency across horizons. Experiments on ten real-world charging stations show consistent improvements over conventional models, reducing MSE and Pinball loss by up to 22.6% for 4‑hour ahead forecasts.

By Chenghan Li, Qingxiang Liu, Yinliang Xu, Yuxuan Liang
More like this →
arXiv AI
Jul 15

BattVAE-GP: Generative Modeling of Long-Horizon Battery Degradation with Uncertainty Quantification

arXiv:2607. 11943v1 Announce Type: cross Abstract: Long-horizon physics-based simulations of battery degradation provide mechanistic insight but remain computationally expensive, limiting their use for dense exploration of operating conditions over extended cycle life.

By Raghvender Raghvender, Mahdi Abid, Ferran Brosa Planella, Charles Delacourt, Arnaud Demorti\`ere
diffusion
More like this →
arXiv Machine Learning
Jun 10

Toward Proactive RF Charging Scheduling: Generative AI for Decision Support

arXiv:2606. 10600v1 Announce Type: cross Abstract: Radio frequency wireless power transfer (RF-WPT) is an enabling technology for supporting uninterrupted communications in future Internet of Things systems by reducing the need for battery replacement and mitigating battery-waste-related issues.

By Amirhossein Azarbahram, Osmel M. Rosabal, David Ernesto Ruiz-Guirola, Melike Erol-Kantarci, Kaibin Huang, Onel L. A. L\'opez
More like this →