An intelligent agent-based computational approach combined with traditional optimization techniques forms a powerful simulation platform to investigate performance of a wholesale electricity market and behaviour of its participants. Modern deregulated wholesale electricity markets consist of centralized auctions as well as decentralized bilateral transactions. An agent-based system is well suited to model the decentralized aspect of modern electricity markets because various market participants can be represented by autonomous agents. Each market participant has its own private goals and it must learn to survive in a dynamic market environment with incomplete information about other participants.Majority of existing agent-based simulation models deal with day-ahead auctions but not bilateral transactions. On the basis of available mathematical modelling details for bilateral transactions, agent-based models that can simulate combination of day-ahead auction and bilateral transactions are categorized into simplified models and proprietary software. Although complete mathematical and implementation details of bilateral transactions are publicly available for simplified models, they only represent bilateral transactions facilitated by brokers or bulletin-boards. By comparison, mathematical details of bilateral transactions’ models used in proprietary software are not publicly available because of commercial value.This thesis provides accurate and in-depth understanding of decentralized bilateral transactions by presenting detailed mathematical modelling that includes: (i) match making for bilateral transactions by a systematic direct-search approach and (ii) bilateral negotiations between participants with incomplete information about each other but capability to learn from interactions.The thesis also facilitates wholesale electricity market simulation including the newly developed model for bilateral energy transactions as well as previously existing models of day-ahead energy auction and financial transmission instruments.
|Date of Award||1 Oct 2014|
- University Of Strathclyde
|Supervisor||Ivana Kockar (Supervisor) & Stephen McArthur (Supervisor)|