A hierarchical multi‑timescale reinforcement learning framework for hybrid energy storage management with aging‑aware constraint adaptation
Mohsen Mohammadbeigi, Alireza Rezazadeh, Mostafa Sedighizadeh · 2026
Read A hierarchical multi‑timescale reinforcement learning framework for hybrid energy storage management with aging‑aware constraint adaptation by Mohsen Mohammadbeigi, Alireza Rezazadeh, Mostafa Sedighizadeh (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13478629.txt.
Description
Mohsen Mohammadbeigi and two colleagues built in 2026 a layered learning controller for battery-supercapacitor hybrid storage. Flat learning designs tangle fast power choices with slow battery ageing, causing shaky training and hard-to-read rewards.
Their design splits control into three layers. A fast safety layer enforces hard limits, a middle layer allocates power by proximal policy optimization, and a slow supervisor tracks battery health with a Gaussian process ageing model. Ageing enters through health-dependent current limits rather than the reward, keeping learning steady. Tests showed charge held within 3 percent of target, slower health decline than flat baselines, and the fewest safety trips under harsh pulsed loads.
Catalog record
| CREATOR | Mohsen Mohammadbeigi, Alireza Rezazadeh, Mostafa Sedighizadeh |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 158 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |