Multi-Objective Optimization and Online Adaptation Methods for Robust Tuning of PSS Parameters

This paper addresses a robust multi-objective optimization approaches for tuning Generator Excitation PSS system parameters of power systems. The objective function which is a composite of different performance indices corresponding to different disturbances and steady-state operating conditions is then posed as a multi-objective nonlinear optimization problem together with parameters from a given set constraint. The nonlinear time solutions for such optimization problem, i.e., the robust parameter set of the Generator Excitation PSS system, are then solved using Sequential Quadratic Programming (SQP) algorithm. Furthermore, an online adaptation of PSS parameters using artificial neural network (ANN) technique in which a pre-calculated optimal parameter sets from the nonlinear optimization solutions corresponding to different loading conditions are then used for training sessions so as to determine the robust parameter sets for the PSS. The proposed approaches have been applied on test system and nonlinear simulation studies have confirmed the robustness of the approaches for all envisaged operating conditions and disturbance scenarios.