The power system is undergoing a transformation towards a more decentralized architecture with the development of local markets and decentralized resources. These markets are fundamental for enabling end-users, including electric vehicle (EV) aggregators, to actively participate in the energy environment. EV aggregators can provide valuable services to both Distribution System Operators (DSOs) and Transmission System Operators (TSOs), such as demand response, frequency regulation, and grid balancing. While many studies have applied various strategic bidding techniques to wholesale markets, there is a notable research gap in exploring their application to smaller environments, such as service markets where EVs operate. In such decentralized frameworks, information is often unavailable in advance, and EV aggregators must frequently adapt their strategies to random scenarios, including fluctuating user demand, charging needs, and grid requirements. Artificial intelligence (AI) and machine learning (ML) offer significant potential in this context. By utilizing AI and ML, it is possible to forecast user behaviour, such as charging patterns and energy needs of electric vehicles and integrate this into the development of strategic bidding techniques tailored for service markets. This project aims at covering this gap by conducting two separate studies. The first step is to investigate how AI and ML algorithms can be employed to forecast user behaviour, improve decision-making, and optimize strategic bidding. The second step concerns the development of an AI-based framework for EV aggregators bidding in service markets. In particular, a case study where a selected bidding technique is implemented to demonstrate its technical and economic potential using AI and/or ML methods. The study wants to assess the impact of AI-driven strategies on the efficiency and profitability of service market participation, examining how EVs can provide ancillary services to DSOs and TSOs while ensuring user satisfaction and grid stability.
