Charging and discharging control of plug-in electric vehicles with uncertainties via robust model predictive control method

With the development of the Plug-in Electric Vehicles (PEVs), it has become a significant part of electric load. The optimal PEVs' charging and discharging scheduling is a vital problem when PEVs connect into the smart grid. In most literatures, the charging and discharging procedure is assumed idealized. In reality, many uncertainties are exist during the procedure, such as the conversion efficiency, temperature and other aspects, but they are not easy to be expressed as a fixed term. In this paper, we use a random but bounded uncertainties to describe the uncertainties. Meanwhile, based on robust model predictive control (RMPC) method, we introduce the disturbance invariant set to solve the PEVs scheduling problem with uncertainties and design a feedback control law to guarantee the feasibility of it. A distributed method to reduce the computation complexity of the uncertain problem. At last, some simulations demonstrate the feasibility of the proposed centralized and distributed methods.