Evolving game theory based decision making systems for NETA power market modelling, analysis and trading strategy development

Yin, Jin (2005) Evolving game theory based decision making systems for NETA power market modelling, analysis and trading strategy development. PhD thesis, University of Glasgow.

Abstract

In this thesis, current work carried out on analyzing the strategic behaviours in electricity trading is first reviewed. An intelligent decision-making and support technique, game theory, is often used in the market practice. Game theory is a discipline concerned with how individuals make decisions when they are partly aware of how their action might affect each other and when each individual might take this into account. Deficiencies and limitations of traditional game theory based methods developed for decision-making in electricity trading are also investigated. This research then explores to discover the impact of intelligent systems based trading strategies in the UK power markets. To model these behaviours and the New Electricity Trading Arrangements (NETA) system of the UK, traditional competitive and cooperative game theory strategies are taken into account in the work reported in this thesis. An improved methodology, “trigger price strategy”, is introduced to simulate power generation companies’ enhanced gaming strategies. Such modelling problem is, however, intractable and hence an extra-numerical search technique, Evolutionary Computation, is employed to solve the game theory based system modelling problem. An encoded Genetic Algorithm based technique is developed to search for an effective model for the complex decision-making process and to help decision-makers evaluate their strategies and bidding parameters.

A novel and effective electricity trading simulation model is thus developed, where its design features are close to the NETA. The model scale is as close as possible to NETA. A complex and more realistic two-sided transaction mechanism with demand fully incorporated is incorporated in this model. These are a world first in this research area.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Subjects: H Social Sciences > HF Commerce; T Technology > TK Electrical engineering. Electronics Nuclear engineering
• Colleges/Schools: College of Science and Engineering > School of Engineering
• Supervisor's Name: Li, Dr. Yun
• Date of Award: 2005
• Depositing User: Elaine Ballantyne
• Unique ID: glathesis:2005-1514
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 17 Feb 2010
• Last Modified: 10 Dec 2012 13:41
• URI: https://theses.gla.ac.uk/id/eprint/1514