Research And Implementation Of Data Visualization System In Power Industry

With the development of smart grid construction,power data shows a rapid growth momentum,bringing new challenges in data analysis and visualization.The development of data analysis and visualization technologies provides a new approach to data visualization in the power industry.Based on the grid investment measurement project,this paper designs a data visualization system for the visualization of power data,and studies the data analysis and visualization problems systematically.The main research contents of this paper are:(1)Radviz data analysis method based on dimensionality reduction and clusteringAiming at the visual display of high-dimensional data in power industry,the large number of dimensions will affect the effect of visual display.This paper proposes a Radviz visualization(PKRadviz)technology based on PCA dimensionality reduction and K-Means clustering,which is realized by PCA dimensionality reduction technology to reduce the overlap rate between data and K-Means cluster analysis to find the relationship between them,in order better to find the relationship and pattern between data.Finally,the visual display of high-dimensional data is realized by Radviz visualization technology,which will not only improves the visualization effect of the traditional Radviz display effectively,but also help to discover the data features or patterns.(2)To improve dimension sorting of radviz visualization technologyFor the visual analysis of high-dimensional data,the traditional Radviz visualization method is greatly influenced by the order of dimension alignment,and different dimension sequences will produce different visualization effects.Research shows that similar dimensions can be placed together to obtain better visualization effects.In this paper,we propose to obtain a better dimension order by calculating the dimensional matrix similarity function.For the solution of the dimensional similar matrix function,the ant colony algorithm,simulated annealing algorithm and simulated annealing-ant colony algorithm are used to calculate it.Through experimental comparison,simulated annealing-ant colony algorithm has better computing performance.By Dunn index evaluation index,The algorithm also gets better data visualization clustering effect.(3)Design and implementation of data visualization system for grid investment measurement and calculationThrough the overall functional design and architecture analysis of the system,the B/S architecture is adopted to realize the separation of front and rear ends.Based on the principle of layered and asynchronous design,the system is divided into four main modules: front-end interface,business logic layer,and permission control layer and the database system.Data visualization technologies are used to display the data and analysis results in a variety of ways in order to see valuable data information and guide the subsequent production work more intuitively.