عنوان

Reinforcement learning based real-time scheduling coastal energy hub Reinforcement learning based real-time scheduling coastal energy hub

Number

T29077

.Language of Text, Soundtrack etc

انگلیسی

Title Proper

Reinforcement learning based real-time scheduling coastal energy hub Reinforcement learning based real-time scheduling coastal energy hub

General Material Designation

Dissertation

First Statement of Responsibility

Hatef Azami

Name of Publisher, Distributor, etc.

Electrical and Computer Engineering

Date of Publication, Distribution, etc.

1402

Specific Material Designation and Extent of Item

146p.

Other Physical Details

cd

Dissertation or thesis details and type of degree

M.S.

Discipline of degree

fulfillment of the requirements

Date of degree

1402/05/02

Text of Note

In this thesis, the emergence of coastal energy hubs is discussed, which has been driven by recent advancements in renewable energy and energy storage technologies. However, scheduling energy production and consumption in these hubs presents significant challenges due to uncertainties associated with energy supply and demand. Existing model-based optimization approaches have been proposed for this purpose, but they have limitations in terms of solution accuracy and computational efficiency, which hinder their practical applications. To overcome these limitations, a model-free, safe deep reinforcement learning (DRL) approach is proposed in this thesis. The proposed approach utilizes primal-dual optimization with the Lagrangian method for optimal scheduling of coastal energy hubs. By leveraging the power of DRL, the proposed approach can learn from experience and adapt to changing environmental conditions. Furthermore, the utilization of primal-dual optimization with the Lagrangian method ensures the safety and stability of the algorithm during the learning process. The proposed approach offers a promising solution for the optimal scheduling of coastal energy hubs, which can enhance the efficiency and sustainability of energy production and consumption in these areas. The findings of the study demonstrate that the proposed approach offers significant benefits in terms of cost reduction and operational constraint satisfaction. The experimental results indicate that the algorithm can effectively reduce the operational costs associated with the scheduling of coastal energy hubs comparison to DDPG and DQN 34% and 38%, respectively.

Text of Note

فاقد چکیده فارسی میباشد

Variant Title

برنامه ریزی کوتاه مدت هاب انرژی ساحی با ستفاده از یادگیری تقویتی

Subject Term

Deep reinforcement learning, Lagrangian multiplier, Floating PV and wind, Real-time scheduling, Episode, Training and testing performance

Subject Term

فاقدکلیدواژه فارسی می باشد.

Entry Element

Azami,

Part of Name Other than Entry Element

Hatef

Relator Code

Producer

Entry Element

Mohammadi-Ivatloo,

Entry Element

Abapour,

Part of Name Other than Entry Element

Behnam

Part of Name Other than Entry Element

Mehdi

Relator Code

Thesis advisor

Relator Code

Thesis advisor

Entry Element

Tabriz

Country

ایران

Agency

Central Library of Tabriz University

Date of Transaction

20230920

Call Number

ارشد پایاننامه TK7871.A9 1402

Electronic name

Hatef Azami

Contact for access assistance

عبادی

e

TL

276903

a

Y