Manifold optimized GAN based EV charging session data generation

doi:10.1016/j.apenergy.2026.127978

Manifold optimized GAN based EV charging session data generation

2026-07-02

0

PRE

AI

DOI：10.1016/j.apenergy.2026.127978

Original

Original request for help

Save

Abstract

En 中文

• A concise and theoretically supported method enhancing Generative AI performance. • Conditional generation of EV charge sessions across year, station, vehicle type, and weekday/weekend settings. • Zero-shot and few-shot generation capability enabling scalability to unseen conditions. • Excellent generalization with data augmentation boosting downstream charge session cost prediction. • Lightweight framework with fewer than 1000 parameters, practical for deployment in energy systems.

Manifold optimized GAN based EV charging session data generation

Abstract

Journal

Organization