Hybrid RBF neural network adaptive optimal power system stabilizer with Tabu search

This paper presents a new systematic to design adaptive optimal power system stabilizer (AOPSS) based on a linear quadratic regulator. Two artificial intelligence programs are applied to design the proposed AOPSS. First, Tabu search is applied to optimize set of the weighting matrices Q and R over their wide range of operating conditions. Second, a radial basis function (RBF) neural network with orthogonal least square learning is used to adapt feedback gains, which correspond with Riccati equation. The proposed method AOPSS is illustrated with application to design of stabilizer for a single generator connected to an infinite bus under various disturbances. The eigenvalue analysis and the time domain simulation results show the effectiveness of the proposed AOPSS to improve damping characteristics and to enhance the system stability over wide range of loading conditions. Moreover, AOPSS gives better performance than conventional stabilizer and adaptive conventional stabilizer.