Integration of artificial neural network in a IEEE 5 bus system

Abstract

The management and the operation of the modern grid is getting more and more complex, especially with the introduction of distributed generation. Grid stability considerations in contemporary power systems are also influenced by the integration of renewable energy, modifications in consumer behaviour, and emerging technology. The study of faults impacts in the distribution network is pertinent and important since it helps to increase the efficiency, safety, and dependability of the power system. It gives utilities the ability to react rapidly to problems, improve maintenance procedures, and prepare for the integration of new technologies—all of which are essential for providing consumers with steady and uninterrupted electricity. In the context of Artificial Neural Networks (ANNs), the Levenberg-Marquardt (LM) algorithm is an extensively utilized optimization method. It is generally used to train feedforward neural networks, especially when those networks have several layers and complicated optimization problems. Introducing a fault in a IEEE 5 bus system help to analyse the changes in the system. The data collected from that phase to phase voltage fault are computed and train in the proposed model and proves the higher efficiency and ability to detect faults or abnormal disturbances. The simulations are done using MATLAB/Simulink.