iDroop: A dynamic droop controller to decouple power grid's steady-state and dynamic performance (1612.05804v1)

Abstract: This paper presents a novel Dynam-i-c Droop (iDroop) control mechanism to perform primary frequency control with gird-connected inverters that improves the network dynamic performance. The work is motivated by the dynamic degradation experienced by the power grid due to the increase in asynchronous inverted-based generation. We show that the widely suggested virtual inertia solution suffers from unbounded noise amplification (infinite $\mathcal H_2$ norm) when measurement noise is considered. This suggests that virtual inertia could potentially further degrade the grid performance once broadly deployed. This motivates the proposed solution in this paper that overcomes the limitations of virtual inertia controllers while sharing the same advantages of traditional droop control. In particular, our iDroop controllers are decentralized, rebalance supply and demand, and provide power sharing. Furthermore, our solution can improve the dynamic performance without affecting the steady state solution. Our algorithm can be incrementally deployed and can be guaranteed to be stable using a decentralized sufficient stability condition on the parameter values. We illustrate several features of our solution using numerical simulations.