Predicting Smart Grid Stability with Deep Learning

Abstract

Smart grid is an advance concept of power system which harmonize electricity and communication in system network. It provides information for the producers, operator and the consumers on real time. There is an extreme demand to efficiently conduct this power supplied to the consumption domains such as household, organizations, industries, and smart cities. For this respect, a smart grid with stable systems is being expended to supply the dynamic power requirements. Predicting smart grid stability is still challenging. Many factors affect the stability of grid one of them is customer and producer participation. Identifying the participants may lead to the smart grid stabilities. In this work, we propose a deep learning model to detect the stability of the smart grid. The results of the proposed model are compared to other popular classifier models used in different studies like Support Vector Machine, Logistic Regression, Decision Tree, Random Forest, Gradient Boosted Trees, Multilayer neural network, Gated Recurrent Units, traditional LSTM and Recurrent Neural Networks and Multidirectional Long Short-Term Memory. The proposed model outperforms the other models with 98.35% accuracy.