Neural Network Predictive Control of UPFC for Enhancing Transient Stability Performance of a Single Machine Infinite Bus System

With an increase in power transfer, transient stability is increasingly becoming important for secure operation. With the advent of FACTS devices, the problem of transient stability is being addressed and efforts are under way to utilize the systems to their full capacity. The focus of this work is a neural network predictive controller for a versatile FACTs device, Unified Power Flow Controller, which can be utilized to improve the transient stability performance of the system. A single machine infinite bus system subjected to a 3-phase short circuit fault is used to demonstrate the superior damping performance of the proposed controller in comparison to the conventional PI controller. Keywords—transient stability, neural network, predictive control, UPFC