Effective identification of a turbogenerator in a SMIB power system using Fuzzy Neural Networks

Abstract

This paper presents modelling and identification of a turbogenerator in a single-machine-infinite-bus (SMIB) power grid utilizing Fuzzy Neural Networks (FuNNs) to construct an online adaptive identifier for the turbogenerator. It is well known that a turbogenerator is a highly nonlinear, fast acting and multivariable system usually connected to a power system. When major power system disturbances occur, protection and control actions are required to stop power system instability and restore the system to a normal state by minimizing the impact of the disturbance. Therefore, effective intelligent techniques are required to model and identify such a complex system. In this paper, a FuNN identifier (FuNNI) of a turbogenerator model is proposed. Computer simulations are carried out to investigate the modelling after deriving the mathematical model of the turbogenerator equipped with a conventional turbine governor and automatic voltage regulator (AVR). Inverse identification scheme is adopted using a multi-input multi-output (MIMO) fuzzy neural network. Empirical results show that the proposed FuNNI is capable of successfully identifying a highly nonlinear turbogenerator system and robust even when the configurations of the plant change due to faults in the power system.