Intelligent AVR and PSS with Adaptive Hybrid Learning Algorithm

The paper presents a step-by-step design methodology of an adaptive neuro-fuzzy inference system (ANFIS) based automatic voltage regulator (AVR) and power system stabilizer (PSS) and also demonstrates its performance in a single-machine-infinite-bus and a multi-machine power system through digital simulation. The design employs a zero and a first order Sugeno fuzzy model, whose parameters are tuned off-line through hybrid learning algorithm. This algorithm is a combination of least square estimator and error backpropagation method. The performance of this ANFIS-based AVR and PSS in damping both local and inter-area oscillation is then compared with conventional fuzzy AVR and PSS performances. It is found that the damping characteristics of both ANFIS-based AVR and PSS are better than the conventional fuzzy AVR and PSS. The effectiveness of the proposed ANFIS-based AVR and PSS in small-signal stability is thus established.