Optimal portfolio selection for generators in the electricity market

With the deregulation of global electricity industries, generators are facing a problem of designing the optimal portfolio, which is made up of a variety of markets and contracts, in a competitive electricity market. Theoretically, the portfolio selection problem can be solved by allocating generation capacities to proper markets and financial contracts, so as to obtain the optimal trade-off between the portfolio return and risk. In this paper, we propose a novel approach which can well solve the generator portfolio selection problem. Considering different planning horizons in practice, the problem can be converted into two sub-problems, which are long term portfolio selection and short term portfolio selection. The mathematical formulations of different asset returns for both the long term and short term portfolio selection have been derived. To model the highly volatile spot market price, a time varying volatility model is introduced. The portfolio selection problem is finally formulated as an optimization problem, which can be solved by the Differential Evolution algorithm. The proposed method is tested with real market data. Cased studies further validate its effectiveness by real market data and demonstrate its promising performance achieved.