基于时序数据压缩的大数据无损编码转换  

Lossless Encoding Conversion of Big Data based on Time Series Data Compression

作　　者：崔赛英 CUI Saiying(School of Artificial Intelligence&Information Engineering,West Yunnan University,Lincang 677000,China)

机构地区：[1]滇西科技师范学院智能与信息工程学院,云南临沧677000

出　　处：《成都工业学院学报》2024年第3期40-44,共5页Journal of Chengdu Technological University

基　　金：云南省教育厅科学研究基金项目(2023J1258)。

摘　　要：针对当下时序数据压缩普遍存在压缩比小、压缩效率低的问题,进行基于时序数据压缩算法的海量大数据无损编码转换研究。该研究分为2部分,首先利用经验模态分解(EMD)算法对时序数据进行分解,分解为有效分量和噪声分量。其次,针对有效分量,利用Huffman算法进行压缩编码转换;针对噪声分量,利用LZ77算法进行压缩编码转换。实验结果表明:与3种传统压缩编码转换算法相比,该算法分别对Haptics和Phoneme数据集进行压缩的均方根失真度为3.854和3.624,压缩比为53.62%和47.85%,由此说明该算法更能够保证在不失真的前提下,以更快的速度完成数据压缩。Aiming at the problems of low compression ratio and low compression efficiency,the lossless encoding conversion of massive big data based on time series data compression algorithm was studied.The research was divided into two parts.Firstly,the empirical mode decomposition(EMD)algorithm was used to decompose the time series data into effective components and noise components.Secondly,for the effective component,the Huffman algorithm was used to compress the encoding conversion.Then LZ77 algorithm was used to compress and encode the noise component.The experimental results show that compared with the three traditional compression encoding conversion algorithms,the RMS distortion of the Haptics and Phoneme data sets compressed by the proposed algorithm is 3.854 and 3.624,and the compression ratio is 53.62%and 47.85%,which indicates that the proposed algorithm can complete the data compression faster without losing the truth.

关 键 词：时序数据 无损压缩算法 EMD算法 HUFFMAN算法 编码转换 

分 类 号：TP241.98[自动化与计算机技术—检测技术与自动化装置]