Big data analysis of smart grid based on back-propagation neural network algorithm in Hadoop environment

Abstract

Smart micro-grid, as an important way for the best use of renewable energy, can help to solve the problem of energy crisis. However, smart micro-grid has limitations to the impact of natural geographical environment conditions, which increases the instability of productions. In this paper, the back-propagation neural network algorithm is used to analyse the data of wind, and the future wind speed is forecasted through the established network model, experimental data is stored and managed in Hadoop framework. The accuracy of predicted results is presented by Relative Error (RE).