Automatic Generation Control (AGC), also known as Load Frequency Control (LFC), operates continuously in modern power systems to maintain power balance among interconnected areas and regulate tie-line power flows. Since electrical energy cannot be stored in large quantities, power generation must match consumption at every instant. This real-time balance is essential for reliable power system management, particularly with the growing integration of renewable energy sources. A typical LFC model includes a turbine, generator, and governor—the latter often exhibiting nonlinearities such as dead band or backlash. These nonlinearities can give rise to self-sustained oscillations or limit cycles (LCs), which degrade the governor’s performance and overall system stability. Even advanced robust PID controllers are often unable to suppress these sustained oscillations. In this work, a digital deadbeat controller is proposed to eliminate such oscillations completely. The proposed LFC strategy operates in four stages. First, the model’s ability to detect frequency deviation (Δf = 0) under a load disturbance (ΔPL) is verified through step response analysis, with the backlash nonlinearity removed from the governor. Next, the LCs are estimated for two benchmark models and the proposed model. Deterministic and Gaussian signals are then applied to suppress the LCs in all five models. Finally, the digital deadbeat controller is implemented for all models in the presence of a stabilizing forcing signal. All simulation studies are carried out using the SIMULINK Toolbox in MATLAB. The results demonstrate that the proposed controller achieves transient-free, ripple-free, and zero steady-state error responses.
Patra et al. (Sun,) studied this question.