Fault prediction in transmission line using GOOSE protocol and Artificial Neural Network

Abstract

To detect and isolate all faulty and worn-out components in the system is the key objective of the detection system. By which the disturbances in the rest of the electrical model are reduced. Thus, the protection system must be reliable. The power communication service object is continually being upgraded by the modernization level of the switch architecture developed in the automation environment of the substation with the advent of recent information technologies. The communication network not only provides a transmission connection between devices, but it also considers the power system's communication network's dependability and efficiency. Many protocols, such as; Intelligent Electronic Devices (IEDs) are employed to offer faster and more reliable communication between control centers. Due to the high volume of traffic in power system communications, many protocols do not offer the required performance with typical switch architectures. This article employs fault detection approaches based on the Generic Object Oriented Substation Event (GOOSE) protocol to overcome this problem. The model is implemented using the MATLAB 2016a program. The classification among faulty and non-faulty transmission lines is made by using the Artificial Neural Network (ANN) classifier. Keywords— Fault Detection, Intelligent Electronic Devices (IED), Generic Object Oriented Substation Event (GOOSE) Protocol, Artificial Neural Network (ANN), MATLAB.