A Consensus-Based Distributed MPC Approach for Batteries Sharing in Group of Buildings

This paper addresses the problem of sharing a private Electrical energy Storage System (ESS) for thermoelectrical energy management in a group of buildings. Each building is equipped with an ESS which can be used partly or completely by other buildings in the same group. This sharing strategy creates coupling constraints among the controllers of each individual building. In this context, an increment proximal minimization method has been employed to manage the energy flows among the various ESS's through a distributed approach. The local controller of each building employs a Model Predictive Control (MPC) which focuses on balancing economic optimization and occupant comfort while fulfilling various local technical constraints. The proposed technique allows local controllers to operate with maximum autonomy and privacy since, at each iteration, a little amount of information is exchanged between each local controller to a centralized coordination unit. Indeed, only the information related to the coupling constraints is required to exchange with the centralized unit. The most significant advantage of the method is to allow the individual building to exploit effectively not only its own ESS but also any excess power and energy capacity of the ESSs of other buildings. The simulation results show the accuracy and efficiency of the proposed method.