Research On Visualization Of Big Data In Smart Grid

With the rapid development of artificial intelligence and big data technology,national grid entered the era of intelligence.The amount of data grows rapidly,how to deal with these data has become a challenge for the development of smart grid,the technology development of related technology and large data visualization has brought new opportunities for the grid data.In the current mainstream big data technology,The Spark platform is superior to the Hadoop platform,workload performance and the high-speed computing power of RDD data sets show great advantages in mass power grid data processing.In this paper,According to the actual demand of power grid data,Zepelin framework and Spark platform are combined.the traditional K-means clustering algorithm is improved,and the improved algorithm is applied to the Spark platform,realize the pretreatment of power grid data;then the traditional Radviz model is improved,and the improved Radviz model is applied to the Zepelin framework.Visualization of power grid data.This paper mainly completed the following three aspects of work: First,the traditional K-means clustering algorithm is studied.According to the shortcomings of the traditional algorithm,an improved K-means clustering algorithm based on clustering efficient function is proposed,and the feasibility of the new efficient function is verified;Second,study the Radviz visualization technology,according to the deficiency of the traditional Radviz model to improve it,put forward a new Radviz model,reduces the coverage between the data,improve the accuracy of the visual results,and the accuracy of the improved Radviz visualization technique is verified by experiment;Third,build the Spark platform and the Zepelin framework,and then combine the two,K-means clustering algorithm is applied to the Spark platform to realize the pretreatment of data grid will be improved,the improved Radviz visualization technology is applied to the Zepelin framework,realize the visualization of data grid,based on grid data visualization graph to verify the accuracy of improved visualization technology.