Intelligent dynamic pricing and integrated demand response for multi-energy systems using Deep Reinforcement Learning

Almannouny, Gaddafi Abbas (2025) Intelligent dynamic pricing and integrated demand response for multi-energy systems using Deep Reinforcement Learning. PhD thesis, University of Glasgow.

Abstract

The increasing penetration of renewable energy sources (RES) and distributed energy systems (DES) presents significant challenges for the power industry, particularly in ensuring grid stability and optimising energy market operations. This thesis investigates the integration of Dynamic Pricing Integrated Demand Response (IDR) into multi-energy systems using Deep Reinforcement Learning (DRL) algorithms to improve efficiency, grid stability, and stakeholder benefits in decentralised energy markets. The first study introduces a dynamic pricing mechanism for electricity and gas systems utilising the Deep Deterministic Policy Gradient (DDPG) algorithm. This mechanism optimises the supply-demand balance, enhances Distribution System Operators (DSOs) profitability, and reduces end-user costs. The second study expands this framework to manage multiple energy carriers— electricity, gas, and heat—through energy hubs (EHs). The DDPG-based IDR strategy promotes cost efficiency and operational flexibility while handling diverse energy demands sustainably. The third study integrates dynamic pricing IDR within a Peer-to-Peer (P2P) energy trading framework for microgrids, employing the Double Actors Regularized Critics (DARC) algorithm. This approach improves renewable energy utilisation, minimises energy deficits, and boosts profitability, outperforming traditional pricing models. The research includes case studies demonstrating the benefits of dynamic pricing and IDR, such as reduced peak loads, increased renewable integration, and enhanced consumer engagement. In conclusion, the thesis lays a foundation for intelligent energy management solutions and suggests future research avenues, including the potential of blockchain technology for P2P trading and advanced consumer behaviour modelling.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Additional Information: Supported by funding from Libyan Government and the Ministry of Higher Education.
• Subjects: T Technology > T Technology (General)
• Colleges/Schools: College of Science and Engineering > School of Engineering
• Funder's Name: Libyan Government and the Ministry of Higher Education
• Supervisor's Name: Yang, Dr. Jin
• Date of Award: 2025
• Depositing User: Theses Team
• Unique ID: glathesis:2025-85367
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 06 Aug 2025 09:04
• Last Modified: 06 Aug 2025 09:06
• Thesis DOI: 10.5525/gla.thesis.85367
• URI: https://theses.gla.ac.uk/id/eprint/85367

Related URLs

• Conference proceeding