A robust adaptive predictive load frequency controller to compensate for model mismatch

The paper describes the design and implementation of a self-tuning load frequency controller (LFC) of a power system based on the adaptive generalized predictive (AGP) controller. The application of predictive control in power system control is considered due to its ability to handle disturbances and un modeled dynamics often encountered in interconnected power system environments. The system dynamics is modeled and linearized around its operating point to represent plant in the controller auto-regressive integrated moving average (CARIMA) form suitable for prediction with the presence of disturbance. Simulation results show that the controller exhibits robustness in handling both problems with proper tuning of its parameters, especially when compared to the normal adaptive control methods of minimum variance and pole assignment.