Designing Robust Strategies for Continuous Trading in Contemporary Power Markets

Abstract

In contemporary energy markets participants interact with each other via brokers that are responsible for the proper energy flow to and from their clients (usually in the form of long-term or short-term contracts). Power TAC is a realistic simulation of a real-life energy market, aiming towards providing a better understanding and modeling of modern energy markets, while boosting research on innovative trading strategies. Power TAC models brokers as software agents, competing against each other in Double Auction environments, in order to increase their client base and market share. Current work discusses such a broker agent architecture, striving to maximize his own profit. Within the context of our analysis, Double Auction markets are treated as microeconomic systems and, based on state-of-the-art price formation strategies, the following policies are designed: an adaptive price formation policy, a policy for forecasting energy consumption that employs Time Series Analysis primitives, and two shout update policies, a rule-based policy that acts rather hastily, and one based on Fuzzy Logic. The results are quite encouraging and will certainly call for future research.