A multi-follower bilevel stochastic programming approach for energy management of combined heat and power micro-grids

Abstract

This paper presents a multi-follower bilevel programming approach to solve the 24-h decision-making problem faced by a combined heat and power (CHP) based micro-grid (MG). The framework contains the interests of two different agents: the MG operator/owner (MGO), who procures the maximization of total profit incurred in attending the forecasted demand of consumers via demand response program (DRP) as well as day-ahead (DA) and real-time (RT) markets participation, and the various CHP owners (CHPOs) who procure the maximization of the profits obtained from the thermal and electrical energy sales. The interaction between the entities is determined in a bilateral contract. Further, to deal with various uncertainties, each level is formulated as a stochastic two-stage problem, where the volatility nature of consumers' loads, RT market price and wind speed uncertainties are modeled using autoregressive moving average (ARMA) technique. In this paper, in order to consider realistic model of the problem, on the contrary to the most CHP-based MG scheduling literature, the network operation constraints such as voltage magnitude of buses and line flow limits are taken into account.