| 41 | 0 | 138 |
| Downloads | Citas | Reads |
A prediction method that integrated fuzzy information granulation and a convolutional neural network and long short-term memory(CNN-LSTM) encoder-decoder structure was proposed to address the problems of complex variable correlations, significant nonlinearity,and frequent local fluctuations in multivariate time series prediction. First, the original time series was preprocessed using triangular fuzzy granulation within a fixed time window, achieving noise reduction and dynamic structure compression. Subsequently, a deep hybrid prediction model composed of a convolutional neural network and a long short-term memory network was constructed to extract local spatial features and model temporal dependencies, enabling high-precision point prediction from series. Finally, the proposed method was validated on two representative multivariate time series datasets, namely the real-time power grid energy demand and the Xi'an air quality index(AQI)datasets. The results show that the MAE, MAPE, and RMSE of the proposed method in energy prediction tasks are 269.637, 0.022, and 374.072, respectively, reflecting reductions of 26.69%, 35.29%, and 20.58% conpared with the CNN-LSTM baseline. In the AQI prediction task, the method likewise achieves superior predictive accuracy, with error reductions of 22.82%, 35.65%, and 28.38%, respectively. These outcomes demonstrate the strong generalization capability and robust stability of the proposed model. The proposed model is suitable for non-stationary and multivariate time series prediction scenarios, offering promising applications for urban energy dispatch and environmental quality monitoring.
[1] WU C C, YEH W C, HSU W D, et al. Prediction of fatty liver disease using machine learning algorithms[J]. Computer Methods and Programs in Biomedicine,2019,170:23-29.
[2]张萌，贺绍鹏，朱文立，等．数据驱动的电工装备市场规模预测架构研究[J]．电力大数据，2022, 25(11):56-62.ZHANG M, HE S P, ZHU W L, et al. Research on data-driven prediction framework of electrical equipment market scale[J]. Power Systems and Big Data,2022,25(11):56-62.
[3]贺小星，花向红，鲁铁定，等．时间跨度对GPS坐标序列噪声模型及速度估计影响分析[J]．国防科技大学学报，2017,39(6):12-18.HE X X,HUA X H,LU T D,et al.Effect of time span on GPS time series noise model and velocity estimation[J]. Journal of National University of Defense Technology,2017,39(6):12-18.
[4] CONTRERAS J,ESPINOLA R,NOGALES F J,et al. ARIMA models to predict next-day electricity prices[J].IEEE Transactions on Power Systems,2003,18(3):1014-1020.
[5] ZHANG G P.Time series forecasting using a hybrid ARIMA and neural network model[J].Neurocomputing,2003,50:159-175.
[6] TOAN T D, TRUONG V.Support vector machine for short-term traffic flow prediction and improvement of its model training using nearest neighbor approach[J]. Transportation Research Record,2021,2675(4):362-373.
[7] TAJEUNA E G,BOUGUESSA M,WANG S R.Modeling regime shifts in multiple time series[J].ACM Transactions on Knowledge Discovery from Data,2023,17(8):1-31.
[8] DAS M, GHOSH S K. SemBnet:a semantic Bayesian network for multivariate prediction of meteorological time series data[J]. Pattern Recognition Letters,2017,93:192-201.
[9] LI S W,HUANG H,LU W.A neural networks based method for multivariate time-series forecasting[J].IEEE Access,2021,9:63915-63924.
[10]WANG X Y, LIU H, DU J Z, et al. A longterm multivariate time series forecasting network combining series decomposition and convolutional neural networks[J].Applied Soft Computing,2023,139:110214.
[11]杨海民，潘志松，白玮．时间序列预测方法综述[J]．计算机科学，2019,46(1):21-28.YANG H M, PAN Z S, BAI W. Review of time series prediction methods[J]. Computer Science,2019,46(1):21-28.
[12] LI Z W,LIU F,YANG W J,et al.A survey of convolutional neural networks:analysis, applications,and prospects[J].IEEE Transactions on Neural Networks and Learning Systems, 2022, 33(12):6999-7019.
[13]管业鹏，苏光耀，盛怡．双向长短期记忆网络的时间序列预测方法[J]．西安电子科技大学学报，2024,51(3):103-112.GUAN Y P,SU G Y,SHENG Y.Time series prediction method based on the bidirectional long short-term memory network[J]. Journal of Xidian University,2024,51(3):103-112.
[14]樊围国，陈珂翰．基于浣熊算法优化双向长短期记忆网络的碳价预测[J]．电力科学与工程，2023,39(7):34-41.FAN W G, CHEN K H. Carbon price prediction of bidirectional long short term memory network optimized by Coati optimization algorithm[J].Electric Power Science and Engineering,2023,39(7):34-41.
[15]徐冬梅，王逸阳，王文川．基于贝叶斯优化算法的长短期记忆神经网络模型年径流预测[J]．水电能源科学，2022,40(12):42-46.XU D M, WANG Y Y, WANG W C. Prediction of annual runoff using long short-term memory neural network model based on Bayesian optimization algorithm[J]. Water Resources and Power,2022,40(12):42-46.
[16] ZADEH L A. Fuzzy sets as a basis for a theory of possibility[J]. Fuzzy Sets and Systems, 1999,100:9-34.
[17]HAO Y D,JIANG S R,YU F S,et al.Linear dynamic fuzzy granule based long-term forecasting model of interval-valued time series[J]. Information Sciences,2022,586:563-595.
[18]左亚尧，王铭锋，洪嘉伟，等．一种基于集成学习的多元时间序列预测方法[J]．小型微型计算机系统，2020,41(12):2475-2479.ZUO Y Y, WANG M F, HONG J W, et al.Prediction method of multivariate time series based on ensemble learning[J]. Journal of Chinese Computer Systems,2020,41(12):2475-2479.
[19]曾千千，曾安，潘丹，等．基于最大信息系数的贝叶斯网络结构学习算法[J]．计算机工程，2017,43(8):225-230.ZENG Q Q,ZENG A,PAN D,et al.Bayesian network structure learning algorithm based on maximum information coefficient[J]. Computer Engineering,2017,43(8):225-230.
[20] CHO K, VAN MERRIENBOER B, GULCEHRE C,et al.Learning phrase representations using RNN encoder-decoder for statistical machine translation[C]//Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing(EMNLP).Stroudsburg,PA,USA:ACL,2014:1724-1734.
Basic Information:
DOI：10.20189/j.cnki.CN/61-1527/E.202505011
China Classification Code:O211.61;TP18
Citation Information:
[1]ZHAO Dan,REN Huorong,REN Jincheng.Granularity Encoder-Decoder Framework-Based Multivariate Time Series Prediction[J].Journal of Rocket Force University of Engineering,2025,39(05):117-126.DOI:10.20189/j.cnki.CN/61-1527/E.202505011.
Fund Information:
西安电子科技大学研究生创新基金项目（YJSJ24001）
2025-10-15
2025-10-15