A Novel Weighted Integration Dynamic Time Regularization and Euclidean Distance Optimization Algorithm for Power Data Mining

Abstract

With the development of large-scale construction of smart grid, the edge terminal equipment of power grid will produce a large number of time series power data with great redundancy, which brings great challenges to the storage of edge side of the equipment. In order to reduce the storage cost of edge side, data mining and weight removal are needed. The traditional data mining technology generally adopts the data mining method based on dynamic time regularity, but the disadvantage is that the mining efficiency is low and the adjacent data with low similarity can not be weighed. Aiming at these problems, this paper proposes an algorithm based on weighted integration dynamic time-regulation and Euclidean distance optimization, which can eliminate data redundancy, achieve data mining and weight removal by calculating the similarity between data. Finally, based on the real sampling data of smart grid, the effect of the proposed data mining technology in edge computing security protection system is analyzed and verified.