9卷 第6期
PetroChina Co Ltd., Beijing, 10000, China
纸质出版：2025-12-30
Scan QR Code
Duan Jian & Huiqiang Lian. (2025). Short-Term Gasoline Sales Forecasting Using an Anomaly-Detecting LSTM Autoencoder Neural Network. 创意经济（英文版）, 9(6), 127-137.
Jian Duan,Lian Huiqiang.Short-Term Gasoline Sales Forecasting Using an Anomaly-Detecting LSTM Autoencoder Neural Network[J].Creative Economy,2025,09(06):127-137.
Duan Jian & Huiqiang Lian. (2025). Short-Term Gasoline Sales Forecasting Using an Anomaly-Detecting LSTM Autoencoder Neural Network. 创意经济（英文版）, 9(6), 127-137. DOI： 10.47297/wspceWSP2516-251925.20250906.
Jian Duan,Lian Huiqiang.Short-Term Gasoline Sales Forecasting Using an Anomaly-Detecting LSTM Autoencoder Neural Network[J].Creative Economy,2025,09(06):127-137. DOI： 10.47297/wspceWSP2516-251925.20250906.
Short-term gasoline sales forecasting often faces significant challenges arising from abrupt events
including major holidays
extreme weather conditions
and fluctuations in international crude oil prices. Such events generate anomalous disruptions in sales data
severely distorting both the trend and amplitude of the underlying time series
complicating model identification and compromising prediction accuracy. Precise identification and mitigation of these anomalies are thus critical for constructing robust predictive models. However
conventional outlier detection methods generally lack suitability for real-time or online learning environments. Additionally
integrating these methods effectively with existing time series forecasting frameworks remains problematic
as traditional approaches frequently overlook structural anomalies characterized by sustained periods of deviation triggered by external events. To address this issue
the current study introduces a novel anomaly detection and forecasting method employing an autoencoder-based Long Short-Term Memory (LSTM) neural network. Initially
the method utilizes a state-space model to decompose the original time series into principal structural components. The decomposed components serve as input for training the anomaly detection neural network
which incorporates a modified Extreme Studentized Deviate (ESD) algorithm proposed by Twitter accounts for seasonal patterns and trends. The proposed framework improves the forecasting model's capability to handle high-variance intervals caused by unforeseen external shocks. Empirical validation results indicate that the developed LSTM model
integrated with the improved anomaly detection mechanism
exhibits superior predictive performance compared to standard LSTM models in the context of short-term gasoline sales forecasting.
Ju L , Zheng Q , Zhang J , et al . Estimating transient vertical hyporheic exchange fluxes from streambed temperatures using LSTM-autoencoder-enhanced physics-informed neural networks [J ] . Journal of Hydrology , 2025 : 133721 .
Akkuş M , Karabatak M , Tekin R . Spindle Autoencoder-CNN hybrid model for cardiac arrhythmia classification [J ] . Computers in Biology and Medicine , 2025 , 195 : 110593 .
Rico J C , Alaeddini A , Faruqui S H A , et al . A Generative Framework for Predictive Modeling of Multiple Chronic Conditions Using Graph Variational Autoencoder and Bandit-Optimized Graph Neural Network [J ] . IEEE Journal of Biomedical and Health Informatics , 2025 .
Andriani S , Galantucci S , Iannacone A , et al . CNN-AutoMIC: Combining convolutional neural network and autoencoder to learn non-linear features for knn-based malware image classification [J ] . Computers & Security , 2025 : 104507 .
Zhang C , Oh S K , Fu Z , et al . Kernel contextual fuzzy rule model based on conditional input space partitioning driven by data reconstruction in autoencoder and randomization-based neural networks [J ] . Knowledge-Based Systems , 2025 : 113679 .
Kumar T , Kaur S , Sharma P , et al . Plant Disease Detection and Classification Using Hybrid Model Based on Convolutional Auto Encoder and Convolutional Neural Network [J ] . Computers , Materials & Continua, 2025 , 83 ( 3 ).
Alsafadi F , Furlong A , Wu X . Predicting critical heat flux with uncertainty quantification and domain generalization using conditional variational autoencoders and deep neural networks [J ] . Annals of Nuclear Energy , 2025 , 220 : 111502 .
Rajan S , Ponnan S . Author Correction: An efficient enhanced stacked auto encoder assisted optimized deep neural network for forecasting Dry Eye Disease [J ] . Scientific Reports , 2025 , 15 : 16008 .
Lei X J , Chen Y L , Pan Y . Multi-Source Data with Laplacian Eigenmaps and Denoising Autoencoder for Predicting Microbe-Disease Associations via Convolutional Neural Network [J ] . Journal of Computer Science and Technology , 2025 , 40 ( 2 ): 588 - 604 .
Zhang T , You T , Liu Z , et al . Small sample pipeline DR defect detection based on smooth variational autoencoder and enhanced detection head faster RCNN [J ] . Applied Intelligence , 2025 , 55 ( 8 ): 1 - 29 .
Gai Y , Liu J . Interpretable unsupervised neural network structure for data clustering via differentiable reconstruction of ONMF and sparse autoencoder [J ] . Neural Networks , 2025 : 107504 .
Zhao H , Min S , Fang J , et al . AI-driven music composition: Melody generation using Recurrent Neural Networks and Variational Autoencoders [J ] . Alexandria Engineering Journal , 2025 , 120 : 258 - 270 .
Mei Z , Bi X , Li D , et al . DHHNN: A Dynamic Hypergraph Hyperbolic Neural Network based on variational autoencoder for multimodal data integration and node classification [J ] . Information Fusion , 2025 : 103016 .
Khan A , Sajid M Z , Khan N A , et al . CAD-Skin: a hybrid convolutional Neural Network–autoencoder framework for precise detection and classification of skin lesions and cancer [J ] . Bioengineering , 2025 , 12 ( 4 ): 326 .
Pajila P J , Robinson Y H , Allimuthu U , et al . Enhanced autoencoder and deep neural network (AE&DNN) method for ddos attack recognition [J ] . Wireless Networks , 2025 , 31 ( 4 ): 3277 - 3295 .
Veeramsetty V , Prabhu K K , Dongari R C , et al . Short-Term Load Forecasting in Distribution Substation Using Autoencoder and Radial Basis Function Neural Networks: A Case Study in India [J ] . Computation , 2025 , 13 ( 3 ): 75 .
Valarmathi P , Suganya Y , Saranya K R , et al . Enhancing parkinson disease detection through feature based deep learning with autoencoders and neural networks [J ] . Scientific Reports , 2025 , 15 ( 1 ): 8624 .
Hu L , Lu Y , Feng Y . Concept Drift Detection Based on Deep Neural Networks and Autoencoders [J ] . Applied Sciences , 2025 , 15 ( 6 ): 3056 .
Deshpande S , Rappel H , Hobbs M , et al . Gaussian process regression+ deep neural network autoencoder for probabilistic surrogate modeling in nonlinear mechanics of solids [J ] . Computer Methods in Applied Mechanics and Engineering , 2025 , 437 : 117790 .
Wang Y , Lei X , Chen Y , et al . Circular RNA-Drug Association Prediction Based on Multi-Scale Convolutional Neural Networks and Adversarial Autoencoders [J ] . International Journal of Molecular Sciences , 2025 , 26 ( 4 ): 1509 .
Ranjan P , Kaushal A , Girdhar A , et al . Revolutionizing hyperspectral image classification for limited labeled data: unifying autoencoder-enhanced GANs with convolutional neural networks and zero-shot learning [J ] . Earth Science Informatics , 2025 , 18 ( 2 ): 1 - 26 .
Ayad A G , El-Gayar M M , Hikal N A , et al . Efficient Real-Time anomaly detection in IoT networks using One-Class autoencoder and deep neural network [J ] . Electronics , 2024 , 14 ( 1 ): 104 .
Veeramsetty V , Dhanush A , Nagapradyullatha A , et al . Power quality disturbances classification using autoencoder and radial basis function neural network [J ] . International Journal of Emerging Electric Power Systems , 2024 , 25 ( 6 ): 817 - 842 .
Geng Z , Li X , Ma B , et al . Improved convolution neural network integrating attention based deep sparse auto encoder for network intrusion detection [J ] . Applied Intelligence , 2025 , 55 ( 2 ): 1 - 17 .
Kuptsov P V , Stankevich N V . Reconstruction of neuromorphic dynamics from a single scalar time series using variational autoencoder and neural network map [J ] . Chaos , Solitons & Fractals, 2025 , 191 : 115818 .
Rathore M S , Harsha S P . Intelligent Fault Detection Scheme for Rolling Bearing Based on Generative Adversarial Network and AutoEncoders Using Convolutional Neural Network [J ] . Journal of Vibration Engineering & Technologies , 2024 , 12 ( 8 ): 8979 - 8991 .
Tiwari S , Gehlot A , Singh R , et al . Design of an iterative method for disease prediction in finger millet leaves using graph networks, dyna networks, autoencoders, and recurrent neural networks [J ] . Results in Engineering , 2024 , 24 : 103301 .
Zhang H , Guo K , Chen Y , et al . Remaining useful-life prediction of lithium battery based on neural-network ensemble via conditional variational autoencoder [J ] . Applied Intelligence , 2025 , 55 ( 1 ): 1 - 15 .
Majikumna K U , Zineddine M , Alaoui A E H . FLVAEGWO‐CNN: Grey Wolf Optimisation‐Based CNN for Classification of Olive Leaf Disease via Focal Loss Variational Autoencoder [J ] . Journal of Phytopathology , 2024 , 172 ( 6 ): e13438 .
Wu J , Yang Y , Lin Z , et al . Weak ultrasonic guided wave signal recognition based on one-dimensional convolutional neural network denoising autoencoder and its application to small defect detection in pipelines [J ] . Measurement , 2025 , 242 : 116234 .
Rassam M A . Autoencoder-Based Neural Network Model for Anomaly Detection in Wireless Body Area Networks [J ] . IoT , 2024 , 5 ( 4 ): 852 - 870 .
Hao T , Zhao L , Du Y , et al . Reduced-Order Model of Coal Seam Gas Extraction Pressure Distribution Based on Deep Neural Networks and Convolutional Autoencoders [J ] . Information , 2024 , 15 ( 11 ): 733 .
Luo X , Liu C , Zhao H . Modeling and spatio-temporal analysis on CO2 emissions in the Guangdong-Hong Kong-Macao greater bay area and surrounding cities based on neural network and autoencoder [J ] . Sustainable Cities and Society , 2024 , 103 : 105254 .
Junlian H , Fei M , Wengxia Y . Optimization Analysis of Clothing Fabric Design Based on Autoencoder and Generative Adversarial Neural Network [J ] . International Journal of High Speed Electronics and Systems , 2024 : 2540068 .
Dutta M , Islam Sujan M R , Mojumdar M U , et al . Rice leaf disease classification—a comparative approach using convolutional neural network (CNN), cascading autoencoder with attention residual U-Net (CAAR-U-Net), and mobileNet-V2 architectures [J ] . Technologies , 2024 , 12 ( 11 ): 214 .
Li X , Zhou W , Yin J , et al . Convolutional variational autoencoder and multi-scale attention convolutional neural network based diagnostics on filament current sensors for mass spectrometers [J ] . Engineering Applications of Artificial Intelligence , 2024 , 138 : 109443 .
Related Articles
Related Author
Related Institution
京公网安备11010802024621