A Decentralized Multi-Agent Online Planning Algorithm and Application on Smart Grid

Zhuo, Wenhao

A Decentralized Multi-Agent Online Planning Algorithm and Application on Smart Grid

Zhuo, Wenhao

Permalink

Publication Type:: Thesis
Issue Date:: 2025

Open Access

Copyright Clearance Process

Recently Added
In Progress
Open Access

This item is open access.

Adobe PDF

Download thesisAdobe PDF (891.69 kB)

View statistics

Full metadata record

Field	Value	Language
dc.contributor.author	Zhuo, Wenhao
dc.date.accessioned	2026-04-10T03:12:30Z
dc.date.available	2026-04-10T03:12:30Z
dc.date.issued	2025
dc.identifier.uri	http://hdl.handle.net/10453/194630
dc.description	University of Technology Sydney. Faculty of Engineering and Information Technology.	en_US.UTF-8
dc.description.abstract	In response to the worldwide emission reduction plan, more distributed renewable resources and energy storage systems are expected to be implemented in the next few decades. Achieving high penetration of renewable energy can be challenging due to the inherent variability and intermittency, and an intelligent reconfiguration is needed. One promising solution is to decompose the power grid into many decentralized and distributed microgrids that have independent power supplies with renewable energy resources. Managing these microgrids can be considered as multi-agent reinforcement learning problems, with objectives of maximizing profit gained by energy trading or minimizing the energy cost. Agents include controllers for distributed renewable resources, energy storage systems, electric vehicle charging stations, and consumers who can influence the power grid. Our research aims to address some of the decision-making problems in smart grid systems with multi-agent reinforcement learning, focusing on decentralized and partially observable environments that reflect uncertainties in renewable forecasting, load demand, and electricity prices. Our main contribution is a novel decentralized planning algorithm that is suitable and easy to implement under the "off-line training, on-line play" framework, which has demonstrated excellence in many state-of-the-art artificial intelligence systems.	en_US.UTF-8
dc.format	Thesis (MAnalytics)
dc.language.iso	en_US	en_US.UTF-8
dc.relation	https://opus.lib.uts.edu.au/bitstream/10453/194630/1/thesis.pdf
dc.rights	info:eu-repo/semantics/openAccess
dc.rights	The author owns the copyright in this thesis including all reproduction and reuse rights for the work. The work may not be altered without the permission of the copyright owner. Attribution is essential when quoting or paraphrasing from this thesis.
dc.rights	© 2025 Wenhao Zhuo
dc.rights	au.edu.uts.lib/cph
dc.title	A Decentralized Multi-Agent Online Planning Algorithm and Application on Smart Grid	en_US.UTF-8
dc.type	Thesis
utslib.copyright.status	open_access	*

Abstract:

In response to the worldwide emission reduction plan, more distributed renewable resources and energy storage systems are expected to be implemented in the next few decades. Achieving high penetration of renewable energy can be challenging due to the inherent variability and intermittency, and an intelligent reconfiguration is needed. One promising solution is to decompose the power grid into many decentralized and distributed microgrids that have independent power supplies with renewable energy resources. Managing these microgrids can be considered as multi-agent reinforcement learning problems, with objectives of maximizing profit gained by energy trading or minimizing the energy cost. Agents include controllers for distributed renewable resources, energy storage systems, electric vehicle charging stations, and consumers who can influence the power grid. Our research aims to address some of the decision-making problems in smart grid systems with multi-agent reinforcement learning, focusing on decentralized and partially observable environments that reflect uncertainties in renewable forecasting, load demand, and electricity prices. Our main contribution is a novel decentralized planning algorithm that is suitable and easy to implement under the "off-line training, on-line play" framework, which has demonstrated excellence in many state-of-the-art artificial intelligence systems.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/194630