A Diffusion–Attention-Enhanced Temporal (DATE-TM) Model: A Multi-Feature-Driven Model for Very-Short-Term Household Load Forecasting

With the proliferation of smart home devices and the ever-increasing demand for household energy management, very-short-term load forecasting (VSTLF) has become imperative for energy usage optimization, cost saving and for sustaining grid stability. Despite recent advancements, VSTLF in the househol...

Full description

Saved in:
Bibliographic Details
Main Authors: Yitao Zhao, Jiahao Li, Chuanxu Chen, Quansheng Guan
Format: Article
Language:English
Published: MDPI AG 2025-01-01
Series:Energies
Subjects:
Online Access:https://www.mdpi.com/1996-1073/18/3/486
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850068306329862144
author Yitao Zhao
Jiahao Li
Chuanxu Chen
Quansheng Guan
author_facet Yitao Zhao
Jiahao Li
Chuanxu Chen
Quansheng Guan
author_sort Yitao Zhao
collection DOAJ
description With the proliferation of smart home devices and the ever-increasing demand for household energy management, very-short-term load forecasting (VSTLF) has become imperative for energy usage optimization, cost saving and for sustaining grid stability. Despite recent advancements, VSTLF in the household scenario still poses challenges. For instance, some characteristics (e.g., high-frequency, noisy and non-stationary) exacerbate the data processing and model training procedures, and the heterogeneity in household consumption patterns causes difficulties for models with the generalization capability. Further, the real-time data processing requirement calls for both the high forecasting accuracy and improved computational efficiency. Thus, we propose a diffusion–attention-enhanced temporal (DATE-TM) model with multi-feature fusion to address the above issues. First, the DATE-TM model could integrate residents’ electricity consumption patterns with climatic factors. Then, it extracts the temporal feature using an encoder and meanwhile models the data uncertainty through a diffusion model. Finally, the decoder, enhanced with the attention mechanism, creates the precise prediction for the household load forecasting. Experimental results reveal that DATE-TM significantly surpasses classical neural networks such as BiLSTM and DeepAR, especially in handling the data uncertainty and long-term dependency.
format Article
id doaj-art-a40ca6dd9d2d4d66bfa9b2a533cb007b
institution DOAJ
issn 1996-1073
language English
publishDate 2025-01-01
publisher MDPI AG
record_format Article
series Energies
spelling doaj-art-a40ca6dd9d2d4d66bfa9b2a533cb007b2025-08-20T02:48:06ZengMDPI AGEnergies1996-10732025-01-0118348610.3390/en18030486A Diffusion–Attention-Enhanced Temporal (DATE-TM) Model: A Multi-Feature-Driven Model for Very-Short-Term Household Load ForecastingYitao Zhao0Jiahao Li1Chuanxu Chen2Quansheng Guan3Yunnan Power Grid Co., Ltd., Kunming 650217, ChinaYunnan Power Grid Co., Ltd., Kunming 650217, ChinaSchool of Electronics and Information, South China University of Technology, Guangzhou 510641, ChinaSchool of Electronics and Information, South China University of Technology, Guangzhou 510641, ChinaWith the proliferation of smart home devices and the ever-increasing demand for household energy management, very-short-term load forecasting (VSTLF) has become imperative for energy usage optimization, cost saving and for sustaining grid stability. Despite recent advancements, VSTLF in the household scenario still poses challenges. For instance, some characteristics (e.g., high-frequency, noisy and non-stationary) exacerbate the data processing and model training procedures, and the heterogeneity in household consumption patterns causes difficulties for models with the generalization capability. Further, the real-time data processing requirement calls for both the high forecasting accuracy and improved computational efficiency. Thus, we propose a diffusion–attention-enhanced temporal (DATE-TM) model with multi-feature fusion to address the above issues. First, the DATE-TM model could integrate residents’ electricity consumption patterns with climatic factors. Then, it extracts the temporal feature using an encoder and meanwhile models the data uncertainty through a diffusion model. Finally, the decoder, enhanced with the attention mechanism, creates the precise prediction for the household load forecasting. Experimental results reveal that DATE-TM significantly surpasses classical neural networks such as BiLSTM and DeepAR, especially in handling the data uncertainty and long-term dependency.https://www.mdpi.com/1996-1073/18/3/486very-short-term load forecastingmulti-feature fusiondiffusion modelattention mechanismenergy management
spellingShingle Yitao Zhao
Jiahao Li
Chuanxu Chen
Quansheng Guan
A Diffusion–Attention-Enhanced Temporal (DATE-TM) Model: A Multi-Feature-Driven Model for Very-Short-Term Household Load Forecasting
Energies
very-short-term load forecasting
multi-feature fusion
diffusion model
attention mechanism
energy management
title A Diffusion–Attention-Enhanced Temporal (DATE-TM) Model: A Multi-Feature-Driven Model for Very-Short-Term Household Load Forecasting
title_full A Diffusion–Attention-Enhanced Temporal (DATE-TM) Model: A Multi-Feature-Driven Model for Very-Short-Term Household Load Forecasting
title_fullStr A Diffusion–Attention-Enhanced Temporal (DATE-TM) Model: A Multi-Feature-Driven Model for Very-Short-Term Household Load Forecasting
title_full_unstemmed A Diffusion–Attention-Enhanced Temporal (DATE-TM) Model: A Multi-Feature-Driven Model for Very-Short-Term Household Load Forecasting
title_short A Diffusion–Attention-Enhanced Temporal (DATE-TM) Model: A Multi-Feature-Driven Model for Very-Short-Term Household Load Forecasting
title_sort diffusion attention enhanced temporal date tm model a multi feature driven model for very short term household load forecasting
topic very-short-term load forecasting
multi-feature fusion
diffusion model
attention mechanism
energy management
url https://www.mdpi.com/1996-1073/18/3/486
work_keys_str_mv AT yitaozhao adiffusionattentionenhancedtemporaldatetmmodelamultifeaturedrivenmodelforveryshorttermhouseholdloadforecasting
AT jiahaoli adiffusionattentionenhancedtemporaldatetmmodelamultifeaturedrivenmodelforveryshorttermhouseholdloadforecasting
AT chuanxuchen adiffusionattentionenhancedtemporaldatetmmodelamultifeaturedrivenmodelforveryshorttermhouseholdloadforecasting
AT quanshengguan adiffusionattentionenhancedtemporaldatetmmodelamultifeaturedrivenmodelforveryshorttermhouseholdloadforecasting
AT yitaozhao diffusionattentionenhancedtemporaldatetmmodelamultifeaturedrivenmodelforveryshorttermhouseholdloadforecasting
AT jiahaoli diffusionattentionenhancedtemporaldatetmmodelamultifeaturedrivenmodelforveryshorttermhouseholdloadforecasting
AT chuanxuchen diffusionattentionenhancedtemporaldatetmmodelamultifeaturedrivenmodelforveryshorttermhouseholdloadforecasting
AT quanshengguan diffusionattentionenhancedtemporaldatetmmodelamultifeaturedrivenmodelforveryshorttermhouseholdloadforecasting