In this article a model predictive control (MPC) based frequency control scheme for energy storage units was derived, focusing on the incorporation of stability constraints based on Lyapunov theory and the concept of passivity.The proposed control schemes, guaranteeing closed-loop stability, are applied on a one-area and two-area power system. For the two-area power system, a coordinated (centralized) control and an uncoordinated (decentralized) control approach is conducted.The stability properties of the different MPC setups were derived, implemented and simulated. Furthermore the corresponding control performance was analyzed.
Index TermsMPC, Nonlinear MPC, Stability, Control Lyapunov Functions, Passivity, Constraints, Power Systems.
I. INTRODUCTION A. MotivationTraditionally, power system operation has essentially been based on the assumption, that electricity is reliably and steadily produced by large power plants, which are fully dispatchable, i.e. controllable and have a high frequency inertia. However, a strong trend towards generation of electrical energy by renewable energy sources (RES), i.e. PV units and wind turbines with no or decoupled rotating masses, exists 1 . How power systems could be adapted in order to accommodate for increasing shares of uncontrolled fluctuating RES as well as power market activities is a highly relevant and still open research question. Options to deal with these challenges might be the expansion of transmission capacities, the extensive integration of storage capacities as well as a better exploitation of the inherent flexibility within the power system. Model predictive control (MPC) as an optimal control scheme for regulating grid frequency receives rising attention due to rapidly growing shares of variable RES and thereby arising challenges for power system operation. The choice of MPC as a control approach is especially motivated by its ability to incorporate operational constraints of power systems, which cannot be handled by conventional P/PI-controllers. It enables the provision of frequency control using a generic power system storage unit, e.g. a battery, with given operational constraints, such as the power ramp rate, power rating and storage capacity. Additionally, the recently growing interest in using MPC for control purposes of power systems emerges due to the availability of faster and cheaper computing resources, as a significant computational effort comes with the use of such a receding horizon control scheme.