Hierarchical Predictive Load Control in Smart Grids

In this paper, a hierarchical iterative control algorithm is proposed to optimally balance grid load (both voltage drop and current load) while meeting consumers power demand. The proposed control architecture is based on the assumption that a certain proportion of the consumers can predict their consumption profile and adapt their control strategies to fulfill power consumption constraints. It is assumed that a simplified model of the power grid and a prediction based on historical data for the consumption profile of passive consumers are available. Based on the consumers' consumption profile predictions, the controller proposes constraints for predictively controlled buildings (PCBs) at critical times, so that these consumers can adapt their consumption strategies in advance. The control algorithm also deals with the case of noncompliant consumers. It has low-communication requirements and consumer information is only shared with the hierarchical controller. The method is verified by simulations of an existing power grid with two PCBs. For this purpose, a grid linearization method and a model predictive controller for a building are suggested.