ANN based impedance trajectory detection approach for loss of excitation protection of synchronous generators connected to transmission lines with SVCs

Abstract

• Computational modeling to evaluate the LOE protection of synchronous generators connected to a transmission line with static var compensator (SVC). • Comparative of traditional LOE protection of with a proposed LOE protection method based on a multilayer perceptron-type artificial neural network (ANN). • Show the effectiveness of the proposed ANN LOE protection to reach faster actuation time than the traditional LOE protection. Flexible Alternating Current Transmission Systems (FACTS) connected to Transmission Lines (TLs) make systems more flexible, and optimize their operational capacity. Despite improvements in TL operational aspects, added FACTS can modify the amplitude, voltage, and current angles of the system wherein they are inserted and consequently the measured impedance, which can affect synchronous machine protections that are based on this feature. Firstly, this paper presents a set of computational simulations to evaluate possible changes in the synchronous machine protection connected to a TL with Static Var Compensator (SVC). Finally, is this context the traditional loss of excitation (LOE) protection is comparing with a proposed approach based on the use of a multilayer perceptron-type Artificial Neural Network (ANN). The simulation results show the effectiveness and robustness of the proposed ANN LOE protection to reach faster actuation time than the traditional LOE protection.