Adaptive Neuro-Fuzzy Technique for Tuning Power System Stabilizer

This paper presents an adaptive neuro-fuzzy technique based on Sugeno first-order fuzzy model for tuning the parameters of a conventional phase lead-lag PSS from on online measurements of the generator loading conditions using adaptive neuro-fuzzy inference system (ANFIS) architecture. Input patterns, of the proposed ANFIS, comprising the generator real power P_e and reactive power Q_e are associated with the corresponding output patterns of the desired PSS parameters K_c and T₁(T₃) over a wide range of operating conditions. Once trained the proposed ANFIS-PSS is capable of providing the PSS parameters K_c and T₁ in real time for any generator loading conditions. Simulation results under small disturbances at different loading conditions have been carried out to assess the effectiveness and robustness of the proposed ANF-PSS in enhancing enhance the small signal stability of the power system