Agent-based learning model for assessing strategic generation investments in electricity markets

The liberalization of electricity markets has significantly changed the perspective of the power generation business. Nowadays, generation companies pursue economic goals due their investment decisions are based on expectations of profitability and the risk of their alternatives. These expectations are difficult to predict because they depend upon various factors that are highly uncertain, including both exogenous uncertainties - such as variations of demand and endogenous uncertainties - such as the behavior of competing generation agents. This paper proposes a numerical tool that financially evaluates investment alternatives of generation companies based on a novel adaptive learning technique that links the generation agents' experiences under the current situation considering their expectations of profitability and risk. In this model, the Agent-based Computational Economics approach has been applied. This method represents generation agents through autonomous and heterogeneous entities pursuing economic goals and interacting through computer models.