Using an improved Neural Network with Bacterial Foraging Optimization algorithm for Load Shedding

Abstract

Making accurate load-shedding decisions helps to reduce losses for customers and the power system. This article proposes an improved Artificial Neural Network (ANN) application using the Bacteria Foraging optimization (BFO) algorithm. The proposed load-shedding model uses ANN to identify load-shedding/non-shedding events combined with optimized load-shedding calculations to ensure system stability within allowable limits. The proposed neural network has been experimented on the IEEE 37-bus system, and its performance is compared with conventional neural networks and those improved by other algorithms. The results show that BFO provides high data identification efficiency with minimal training time, making it a potential solution for load-shedding forecasting.