Enhancing UPFC compensation performance in simultaneous power and voltage control of the system with self-tuning PI controllers based on ANFIS design

Ensuring stability and effective oscillation damping in modern power systems remains a critical challenge, particularly under severe disturbances such as load variations or faults. Conventional stabilizers, such as the Power System Stabilizer (PSS), often exhibit limited performance across a wide range of operating conditions. To address these limitations, this paper presents an online adaptive PI–ANFIS controller for the Unified Power Flow Controller (UPFC), designed to simultaneously regulate line power flow, DC-link voltage, and generator terminal voltage. The proposed structure employs parallel ANFIS networks combined with PI controllers, where the PI gains are continuously adapted using a Recursive Least Squares (RLS)-driven learning mechanism, enabling real-time adjustment to nonlinear dynamics and parameter uncertainties. The effectiveness of the controller is validated through comprehensive simulations in MATLAB/Simulink as well as hardware-in-the-loop (HIL) experiments on a Speedgoat real-time target, ensuring both theoretical and practical feasibility. Performance is assessed under nominal and heavy-load conditions, including the most demanding Scenario 4 with 10 % power variation. Comparative evaluations against classical PI, metaheuristic-optimized controllers, and recent advanced designs (FPA–PID, O-FOPID, and NF-3DOF-PDN) demonstrate that the proposed controller achieves superior transient and steady-state responses. In particular, the results confirm significant improvements in settling time (up to 28 % faster), overshoot reduction (over 40 %), and minimized steady-state error (<0.002 pu), while maintaining robustness against parameter variations. These findings establish the proposed online PI–ANFIS controller as a robust, adaptive, and real-time capable solution for enhancing UPFC performance and ensuring resilient power system operation under challenging scenarios.