Resposta à demanda via precificação em tempo real para controle da demanda de energia

Abstract

The energy pricing in a demand response program aims to define price signals at each instant in order to stimulate the shift from the peak energy consumption to the valleys. From a demand curve target drawn a priori, we propose an optimal policy of closed loop control, that prices the energy in real-time by penalizing deviations of the demand curve with respect to this trajectory of power dispatch. The optimal policy is obtained by a stochastic dynamic programming with a daily horizon, assuming that the controls are the kWh tariffs and the states are the loads demanded by the electrical system. The approach of the quadratic cost is used to penalize large deviations from the curve to the desired trajectory. In order to verify the consistency of this mathematical formulation, we derive some properties of the model and analyze its adherence to reality. To illustrate its applicability, a numerical example is presented and we show an optimal policy and simulations of load curves which evolve according to this policy. At the end, we offer a detailed discussion of the formulation and paths for future research.