Optimal Distributed Energy Storage Management Using Relaxed Dantzig-Wolfe Decomposition

摘要

This paper focuses on optimal management of energy storage systems in distribution grids. Performance and operation objectives are formalized in an optimal power flow formulation, where charging and discharging inefficiencies of the batteries are accounted for. Binary optimization variables are utilized to avoid meaningless solutions where a battery is required to simultaneously charge and discharge. Leveraging linearized AC power flow models, the resultant optimization becomes a mixed-integer linear program. A decentralized scheme is then devised based on the Dantzig-Wolfe decomposition algorithm. The efficacy of the developed algorithm is demonstrated using an IEEE distribution feeder equipped with storage units and photovoltaics.