Nonlinear Model Based Control with prior knowledge Based Learning

The ever increasing demand and restriction on having additional new infrastructure, forces the existing power system network to work at its maximum possible limits. In order to increase the power transfer through given infrastructure, the use of FACTS devices is common and very well known. Here a model predictive control based TCSC controller is used for improving the transient stability response of a single machine infinite bus system (SMIB). The system model is used to formulate a quadratic optimization problem to help find out the required capacitive compensation from series connected TCSC in a given line. Considering the nonlinear and hybrid nature of power system a nonlinear model predictive controller (NMPC) is designed. The major contribution of this paper is the use of NMPC strategy for TCSC controller to enhance transient stability of SMIB system considering a detail system model with 10 variables. Since system model considered here is nonlinear in nature, NMPC allows the use of a nonlinear model instead of using a linear model for prediction. Also, NMPC helps to increase the region of attraction, which allows system to come back to healthy conditions even after the fault is removed beyond critical clearing time. For reasons mentioned above, the method suggested here appears to have an edge over other methods employing linear model and controllers for SMIB system. The concept and theory is supported by simulation on SMIB system with nonlinear model predictive based TCSC controller.