Adaptive DOCR coordination in loop distribution system with distributed generation using firefly algorithm-artificial neural network

Abstract

The addition of Distributed Generation (DG) to the power system provides some of the impact of changes on the distribution network. With the addition of DG, it is important to ensure a fast and reliable protection system to avoid accidental disconnection of DG when there is disruption to the distribution network. Another impact of the addition of DG is that protection on the system needs to be coordinated again. In this research, it is proposed coordination of protection which is adaptive and optimal used Firefly Algorithm (FA) and Artificial Neural Network (ANN) to obtain optimal coordination. This study is tested on a modified IEEE 9 bus loop system with the addition of DG. Because the direction of current flowing from different directions so that required coordination of directional over current relay protection (DOCR). Optimization is tested in four different combinations of conditions. Optimization using firefly algorithm will get the value of Time Dial Setting (TDS), Pickup Current (IP) and total of the fastest operation time. Backpropagation algorithm used in ANN training process. The training process uses the input of ISC max taken based on the combination of generation, the fault location, and the type of fault. The TDS and IP values of FA optimization results are used as ANN training targets. After testing, the results obtained in accordance with the target data. The results of both method have been proved by the ETAP simulation which shows that the FA-ANN is a suitable method to model the adaptive and optimal relay coordination system.