Abstract:  Due to the use of power flow equation in tradition voltage control, the voltages may lose its stability in the transition process form the failure to the stable operating point. It is necessary to consider dynamic model of power system in voltage study. This paper presents a nonlinear voltage predictive control method based on dynamic reduced model. In order to decrease the optimal calculating time, the improved empirical Gramian balance reduction method based on the features of power system is applied to reduce the order of power system nonlinear dynamic model. In order to improve the accuracy and numerical stability of predictive model, this paper uses a 4-order convergent Adams method instead of Euler method to predictive state values, and develops a multi-step prediction and rolling optimization model based on reduced model. In addition, the warm start technique and a small iterative times limit Nmax are used to decrease the iterative times in the solving process. The New England 10-generator and 39-bus power system is simulated to test the performance of the proposed method. The simulation results show that this method can not only increase the predictive model numerical stability and decrease the optimal-timing greatly, but also respond the predictable dynamic change of the system in advance to keep voltage stable.

Key words:  power system, voltage control, dynamic model,  model predictive control(MPC), model reduction