A hierarchical resilient tri-level energy management for islanded multi-energy system with risk-aware learning-in-the-loop RL algorithm

Coordinated energy management is crucial for the effective operation of integrated multi-energy systems (IMESs), which integrate water, heat, and electricity networks. The primary objectives of this coordination include reducing operational costs, minimizing hazards, and ensuring a reliable supply of energy. Traditional methodologies tend to address these objectives within a singular optimization framework, which often constrains their capacity to accommodate the distinct operational requirements associated with economic scheduling, vulnerability assessment, and service restoration. This study investigates the potential for enhancing system performance by dispersing these objectives into three interconnected decision-making tiers and coordinating their outputs throughout the operational phase. At the highest level, Particle Swarm Optimization (PSO) is utilized to facilitate operational decisions that improve energy supply and service restoration during disruptions. The intermediate level employs a sensitivity-based risk assessment to pinpoint vulnerable system components and evaluate their effects on operational performance. At the foundational level, resources pertaining to power, heat, water, storage, and energy conversion are scheduled through Mixed-Integer Linear Programming (MILP), which integrates resilience and risk outcomes as operational constraints and decision-making priorities. Furthermore, a Reinforcement Learning (RL) agent manages real-time operations by coordinating the outputs from these three levels, dynamically adjusting operational choices based on prevailing system conditions. This hierarchical flow of information enables resilience requirements and vulnerability assessments to inform economic scheduling without conflating all objectives into a single optimization problem. Simulations conducted on an IMES reveal that the proposed framework enhances energy supply during disruptions by 35.14% and reduces daily operating costs by 13.79% when compared to traditional benchmark approaches. These findings suggest that the proposed structure improves supply continuity and economic efficiency while effectively addressing operational vulnerabilities and network limitations.

Authors

Institutions

Publication Details

Journal
Scientific Reports
Published
2026-11-18
DOI
https://doi.org/10.1038/s41598-026-77395-8
Primary Topic
Integrated Energy Systems Optimization
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

A hierarchical resilient tri-level energy management for islanded multi-energy system with risk-aware learning-in-the-loop RL algorithm

Abbas Zare Ghaleh Seyyedi, Wenwu Tan, Ali Hashim Alwan, Hongyi Liao
Scientific Reports
Integrated Energy Systems Optimization
article

A hierarchical resilient tri-level energy management for islanded multi-energy system with risk-aware learning-in-the-loop RL algorithm

Abbas Zare Ghaleh Seyyedi, Wenwu Tan, Ali Hashim Alwan, Hongyi Liao
article en

Abstract

Coordinated energy management is crucial for the effective operation of integrated multi-energy systems (IMESs), which integrate water, heat, and electricity networks. The primary objectives of this coordination include reducing operational costs, minimizing hazards, and ensuring a reliable supply of energy. Traditional methodologies tend to address these objectives within a singular optimization framework, which often constrains their capacity to accommodate the distinct operational requirements associated with economic scheduling, vulnerability assessment, and service restoration. This study investigates the potential for enhancing system performance by dispersing these objectives into three interconnected decision-making tiers and coordinating their outputs throughout the operational phase. At the highest level, Particle Swarm Optimization (PSO) is utilized to facilitate operational decisions that improve energy supply and service restoration during disruptions. The intermediate level employs a sensitivity-based risk assessment to pinpoint vulnerable system components and evaluate their effects on operational performance. At the foundational level, resources pertaining to power, heat, water, storage, and energy conversion are scheduled through Mixed-Integer Linear Programming (MILP), which integrates resilience and risk outcomes as operational constraints and decision-making priorities. Furthermore, a Reinforcement Learning (RL) agent manages real-time operations by coordinating the outputs from these three levels, dynamically adjusting operational choices based on prevailing system conditions. This hierarchical flow of information enables resilience requirements and vulnerability assessments to inform economic scheduling without conflating all objectives into a single optimization problem. Simulations conducted on an IMES reveal that the proposed framework enhances energy supply during disruptions by 35.14% and reduces daily operating costs by 13.79% when compared to traditional benchmark approaches. These findings suggest that the proposed structure improves supply continuity and economic efficiency while effectively addressing operational vulnerabilities and network limitations.

Scientific Reports
Jiangxi University of Technology (CN), University of Misan (IQ), Xinyu University (CN)
Affordable and clean energy, Industry, innovation and infrastructure
Integrated Energy Systems Optimization
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon