Research On The Feature Analysis Of Electric Power Big Data Based On Hadoop

The world has been gone in the age of big data.2013 was the year of China big data. With the fully implement of smart power grids、Internet of Things and cloud computing, the power big data has emerged.Power big data has determined the economic and society of development of human socity by its special status.How to make better use of power big data for production and living for the people has become a key.Based on clustering algorithm of the power load data, this paper proposes a load curve clustering algorithm which based on cloud computing platform.This clustering algorithm can be better for the production and living of the society.This paper analyzes how to get more useful information from power data under the age of power big data.Firstly this paper analysis and research the environment of cloud computing, Hadoop.That has provided the safeguard for the implementation of data mining technology.Then according to the data mining technology,in-depth analysis of the data preprocessing and clustering algorithm in process of data mining.Under the analysis and research of different principles of clustering algorithm and its application in the field of power load curve, this paper provided a more efficient DWT hierarchical clustering algorithm which based on cloud computing.Finally the paper attempts to apply the actual power data to DWT hierarchical clustering algorithm which based on Hadoop calculation platform.Through the analysis between the clustering result and the actual data, this paper vertified the scheme’s effectiveness and efficiency for applied in the electric power big data.