An Efficient Electricity Generation Forecasting System Using Artificial Neural Network Approach with Big Data

The fact that the United States (U.S.) enjoys a large geographical diversity among states with enormous amount of power consumption makes it challenging to centralize a power management system that can control the power generation and regulate the consumption. Due to lack of centralized control, there is a large imbalance in the ratio of power consumption/power generation from one state to the next. This imbalance results in wasting of large quantities of power generated in states where generation exceeds consumption significantly, whereas other states are suffering from insufficient amount of power generation. Power generation is in direct correlation with the amount of resources used to generate the power. In this paper, we propose a power generation forecasting scheme that could predict the amount of power required at a rate closer to the power consumption. The proposed scheme uses Big Data analytics to process power management data collected in the past 20 years for each state. It then uses a Neural Network (NN) model to train the system for prediction of future power generation based on the collected data. Simulation shows that the proposed scheme can predict power generation close to 99% of the actual usage.