A stable finite horizon model predictive control for power system voltage collapse prevention

This paper proposes a finite horizon model predictive control (MPC) method for online voltage collapse prevention. With a hybrid differential-algebraic equations (DAE) model, a bulk electric power system is studied including both continuous dynamics and certain switching behavior critical to voltage instability. Two major contributing factors for long term instability are then discussed to show how such switching mechanisms could lead to unstable dynamic behavior. Based on this model, a safety set concept constructed from a projection of algebraic variables is proposed. Terminal inequality constraints defined by this safety set are adopted to establish convergence properties for the proposed finite horizon MPC algorithm. Constructing this safety set, which is described only by algebraic variables, does not require detailed dynamic state estimation which is not yet available for large-scale power systems. A 10-bus benchmark case for studying voltage collapse is used to illustrate the performance of the control method.