A SAFE reinforcement learning approach for low-carbon community energy cooperative management considering electric vehicle charging satisfaction

Shouxiang Wang

doi:10.1016/j.seta.2026.105143

A SAFE reinforcement learning approach for low-carbon community energy cooperative management considering electric vehicle charging satisfaction

2026-06-17

0

PRE

AI

王守相 (Shouxiang Wang)

F

Fayang Tian

X

Xuan Zhang

赵

赵倩宇 (Qianyu Zhao) *

L

Luyang Guo

H

Hainan Zhu

DOI：10.1016/j.seta.2026.105143

Original

Original request for help

Save

Abstract

En 中文

• Safe reinforcement learning approach for low-carbon community energy management. • EV-community cooperative framework with electric vehicle charging satisfaction guarantee. • Retrospective constraint mapping and safety layer projection ensure terminal SOC constraint satisfaction. • Adaptive augmented Lagrangian method guides RL agents to balance constraints with rewards. • Validation demonstrates multi-objective optimization across economic efficiency, carbon reduction, and peak load reduction.

Keywords:

Community energy management system

Electric vehicle charging management

Safe reinforcement learning

Constrained Markov decision process

Augmented Lagrangian method

A SAFE reinforcement learning approach for low-carbon community energy cooperative management considering electric vehicle charging satisfaction

Abstract

Journal

Organization

Cited Papers

Citing Papers