A Dimensional Reduction Visualization Method for Large Data of Power Grid Based on t-SNE Algorithms

The large data of power grid includes the data generated or needed in the process of planning, operation, management, and operation of distribution power system, including power generation data, transmission data, distribution data, power consumption data, and external data (such as economic, social and meteorological data). Data dimension is high and visualization is difficult. In this paper, a dimensionality reduction visualization method based on the t-SNE algorithm for large data of power grid is proposed. The dimensionality reduction of large data of high-dimensional power grid is processed on the premise of maintaining local geometric characteristics. Because the t-SNE algorithm has the ability to process non-linear and high-dimensional data, it can meet the needs of multi-type data fusion analysis and processing in a smart grid. In this paper, the 96-dimensional load data in the Shanghai area is used for visualization simulation, and good results are obtained, which provides a novel slotion for the visualization of large data in the high-dimensional power grid.

[1]  Barbara Hammer,et al.  Parametric nonlinear dimensionality reduction using kernel t-SNE , 2015, Neurocomputing.

[2]  Laurens van der Maaten,et al.  Accelerating t-SNE using tree-based algorithms , 2014, J. Mach. Learn. Res..

[3]  Xia Mao,et al.  Facial Expression Recognition Based on t-SNE and AdaboostM2 , 2013, 2013 IEEE International Conference on Green Computing and Communications and IEEE Internet of Things and IEEE Cyber, Physical and Social Computing.

[4]  Geoffrey C. Fox,et al.  High Performance Dimension Reduction and Visualization for Large High-Dimensional Data Analysis , 2010, 2010 10th IEEE/ACM International Conference on Cluster, Cloud and Grid Computing.

[5]  Xianchao Zhu,et al.  Visualization of disease relationships by multiple maps t-SNE regularization based on Nesterov accelerated gradient , 2017, 2017 IEEE International Conference on Bioinformatics and Biomedicine (BIBM).

[6]  Barbara Hammer,et al.  Linear basis-function t-SNE for fast nonlinear dimensionality reduction , 2012, The 2012 International Joint Conference on Neural Networks (IJCNN).

[7]  John T. Stasko,et al.  An interactive visual testbed system for dimension reduction and clustering of large-scale high-dimensional data , 2013, Electronic Imaging.

[8]  Xianchao Zhu,et al.  Visualization of non-metric relationships by adaptive learning multiple maps t-SNE regularization , 2017, 2017 IEEE International Conference on Big Data (Big Data).

[9]  Weiguo Fan,et al.  Trace-Oriented Feature Analysis for Large-Scale Text Data Dimension Reduction , 2011, IEEE Transactions on Knowledge and Data Engineering.

[10]  Gangyao Kuang,et al.  Deep supervised t-SNE for SAR target recognition , 2017, 2017 2nd International Conference on Frontiers of Sensors Technologies (ICFST).

[11]  Kim-Anh Lê Cao,et al.  Independent Principal Component Analysis for biologically meaningful dimension reduction of large biological data sets , 2012, BMC Bioinformatics.