Power Demand Forecasting Using Long Short-Term Memory Neural Network based Smart Grid

Saving energy is still a controversial and promising topic nowadays. Power demand forecasting is an interesting prospect which has started gaining a lot of attention recently in the smart grid. Following various algorithms and models, plenty of predicting methodologies have been proposed in international research. With the rapid development of AI technology, several latest approaches based on artificial neural networks combining with other techniques have been addressed as well. This paper proposes RNN-LSTM, a power demand forecasting model based on Long Short-Term Memory (LSTM) neural network. We used energy data from a real system to test and implement our proposed scheme with some vital features of a predicting power system. Our study also compared with other algorithms such as the linear regression model to show the higher accuracy and performance of our proposed algorithm.