电子科技2025，Vol.38Issue(11)：53-60,8.DOI:10.16180/j.cnki.issn1007-7820.2025.11.007
基于自编码器的神经网络聚合传递时间序列压缩算法
Autoencoder-Based Neural Network Aggregated Transmission Algorithm for Time Series Compression
摘要
Abstract
In view of the problems that multi-dimensional time series occupy a large amount of storage resources due to the increasing demand of time series data collection and transmission by intelligent machines,internet of things devices and sensors,an autoencoder-based time series compression algorithm is proposed in this study.The encoder and decoder are constructed based on the LSTM(Long Short-Term Memory)neural network,and the encoder reduces the dimension of the input data to achieve compression.In order to reduce the data loss of multi-dimensional time se-ries and enhance the accuracy of data reconstruction at the decoder end,the compression amount of the compressed rep-resentation is adjusted through the multi-layer neural network aggregation structure.The proposed compression algo-rithm is evaluated experimentally on a public time series data set,and the results show that the compression rate of 0.15 can be achieved within a maximum allowable error of 0.2.关键词
自编码器/神经网络/时间序列/数据压缩/深度学习/LSTM/压缩率/数据降维Key words
autoencoder/neural network/time series/data compression/deep learning/LSTM/compression ratio/data dimensionality reduction分类
计算机与自动化引用本文复制引用
李易霖,王成群..基于自编码器的神经网络聚合传递时间序列压缩算法[J].电子科技,2025,38(11):53-60,8.基金项目
浙江省重点研发计划(2021C01047) Key R&D Program of Zhejiang(2021C01047) （2021C01047）
