Design of model predictive controllers for adaptive damping of inter-area oscillations

An adaptive damping controller design method by integrating online recursive closed-loop subspace model identification (SMI) with model predictive control theory is proposed in this paper. A reduced order state-space model which contains dominant low frequency oscillation modes is firstly identified online by using an closed-loop SMI algorithm. Then, an infinite horizon closed-loop optimal control is achieved based on model prediction which uses the current state of power system as the initial state. At each control step, the identified model is updated by using the new coming measurements and the optimal control action is solved again. Periodical online model and control updating identification and optimal control enables the proposed controller to adapt to operating condition variations and system parameter uncertainties. It is more robust than offline identification based damping controllers which could suffer from performance degradation under time varying and uncertain conditions. Simulation results demonstrate the effectiveness and robustness of the proposed controller in damping inter-area low frequency oscillations. The abilities to coordinate with power system stabilizers (PSSs) and other similar controllers in multi-machine power systems are also illustrated.

► An adaptive damping controller (ROPEIV-MPC) is designed by integrating SMI with MPC theory.
► ROPEIV-MOESP algorithm fits for closed-loop model identification and has better performances.
► The ROPEIV-MPC uses wide-area feedback signals as inputs and focuses on damping inter-area oscillations.
► The ROPEIV-MPC is robust to operating condition variations and system parameter uncertainties.
► Online model and control updating makes the ROPEIV-MPC can coordinate with other controllers.