An evolutionary game approach to analyzing bidding strategies in electricity markets with elastic demand

摘要

In this paper we propose an evolutionary imperfect information game approach to analyzing bidding strategies in electricity markets with price-elastic demand. In previous research, opponent generation companies' (GENCOs') bidding strategies were assumed to be fixed or subject to a fixed probability distribution. In contrast, the adaptive and learning agents in the presented model can dynamically update their beliefs about opponents' bidding strategies during the simulation. GENCOs are represented as different species in the coevolutionary algorithm to search the equilibrium. By modeling the evolutionary gaming behavior of GENCOs, the simulation can capture the dynamics of GENCOs' strategy change. This is important for analyzing transitory behavior of agents in the market in addition to the long-run equilibrium state. Simulations show that due to the adaptive learning, the bidding evolution is different from the one in the traditional game.