This paper presents a real time strategy for optimal power flow in presence of storage devices and wind turbine driven by Doubly Fed Induction Generators. These elements work in cooperation defining a dynamic bus where the generated power is subject to temporal constraints, which establish a coupling between traditional power flow problems related to consecutive time periods; further the uncertainty in wind power generation forecasts requires a continuous update of the planned power profiles, in order to guarantee a dynamic equilibrium among demand and supply. Model predictive control is used for this purpose, considering the dynamic equations of the storage and the wind turbine rotor as prediction models. A proper target function is introduced in order to find a trade-off between the need of minimizing generation costs and the excursions of the storage state of charge and the wind turbine angular speed from reference states. In the case study under consideration storage, wind turbines and a traditional synchronous generator are operated by the Transmission System Operator in the form of a Virtual Power Plant working as slack bus to cover network losses. The proposed approach is validated on simulation basis.