Predictor-Based Load Frequency Control for Large-Scale Networked Control Power Systems

This paper investigates the load frequency control scheme for the large-scale networked power system with large delays, where the system states are unmeasurable. A matrix dimension transformation method is proposed to solve the problem of matrix dimension mismatch caused by the complex structure of the system in the controller design process. Two formal results are demonstrated. The first result studies the exponential stability using the backstepping-based partial differential equation (PDE) method in the continuous-time control framework. The second result examines the exponential stability based on the the reduction-based ordinary differential equation (ODE) method in the framework of sampled-data control. A two-sided mode-dependent loop-based Lyapunov-Krasovskii functional is constructed so that the information of sampled-data intervals can be utilized more sufficiently. Meanwhile, with the help of proposed method, the design algorithms of controllers are given base on two predictor methods, respectively. And
$H_\infty$
performance indexes reflecting the system robustness under load-frequency control are given. The validity of the proposed methodology is verified in the simulation section using a three-area interconnected large-scale power system.