Distributed Model Predictive Control With Application To Load Frequency Control System

Load frequency control (LFC) is one of the most profitable services of power systems. It can realize the balance between active power and load power. The main goal of the LFC problem is to maintain zero steadystate errors for frequency deviation and good tracking of load demands in a power system. As the electric power system has become increasingly large, the power systems gradually realize interconnected operation. Distributed model predictive control is suitable for large-scale interconnected power system. In this algorithm, the system will be divided into several subsystems. Each subsystem has independent controller. The algorithm can not noly reduce the computation burden, but also meet the requirements of the entire system. In practical power system LFC, the existence of the generation rate constraints (GRC) and the valve position limit in the governor would influence the dynamic responses of the system significantly and lead to larger overshoot and longer settling time. Then, a distributed model predictive control based on T-S model is presented to control a four-area hydro-thermal interconnected power system to solve the nonlinear problems. Considering the system uncertainty according to power system parameter variation, a robust distributed model predictive control is presented. The problem of the power system is formulated as convex optimization problem with linear matrix inequalities (LMI) that can be solved efficiently. It satisfys a robust performance objective for each subsystem. Simulation results demonstrate the effectiveness of the proposed control algorithm.