Model predictive control for load frequency control of an interconnected power system.
Igbineweka, Ernest Uyioghosa.
MetadataShow full item record
Reliable load frequency control (LFC) is of importance in modern power system generation, transmission and distribution, it has been the basis of research on advanced control theory and application in recent years. In LFC scheme, local load disturbance, inter-area ties power fluctuation, frequency deviation, generation rate constraints (GRC), and governor dead band (GDB) are the major nonlinear factors on the control scheme that affect the dynamic response of the system to a large extent. Over the years, many methods have been designed for LFC problem of which model predictive controller (MPC) stands out due to its advantages. MPC is a control approach that simulates the feature behaviour of a system it controls and based on the result of the simulation attempt to find a control output such that the simulated system behaves optimally. When applied to LFC it copes with the perturbation. In this dissertation, robust distributed model predictive control (RDMPC) is developed as a controller scheme for LFC and is compared with a proportional integral derivative (PID) controller using MATLAB/Simulink for two-area and three-area hydro-thermal interconnected power system. From the simulation result, RDMPC significantly shows robustness over PID when compared in frequency deviation and area control error. It is observed that RDMPC still lags, from system varying dynamics and uncertainty despite its robustness over PID, hence an adaptive model predictive control (AMPC) is developed to improve on the performance of RDMPC. In order to evaluate the efficacy of this proposed controller, robustness and comparative analysis is performed using MATLAB/Simulink between the conventional PID, RDMPC, and AMPC with respect to performance indices such as settling time, undershoot and peak overshoot when subjected to frequency deviation, tie-line active power deviation, and area control error. Also, the dynamic response of the hydrothermal systems is analysed and compared in the presence of nonlinear constraints such as generator rate constraint (GRC) and governor dead band (GDB). The result from the simulation tests shows that AMPC has a better dynamic response when compared with PID, and RDMPC with a significant improvement.