Meteorology-sensitive load forecasting for urban sub-sector aggregators based on a PSO-LSTM-RF hybrid model

Read full text Summarise with VIRAT
Authors: Sen Lin, 琦芬 李, Liting Zhang, Fanyue Qian, Yongwen Yang
Cited by: 0
References: 6

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Abstract To address the limitations of single-model approaches and inadequate exploitation of industry-specific heterogeneity in urban sub-sector load forecasting, this study proposes a PSO-LSTM-RF hybrid framework for meteorology-sensitive power load prediction and correlation analysis. Using half-hourly load data from four major industry aggregators and corresponding meteorological observations, we categorize industry loads into three groups, and employ mutual information entropy and Pearson correlation coefficients to quantify nonlinear / linear dependencies and evaluate model performance. Results demonstrate that the PSO-LSTM-RF model outperforms individual models consistently, revealing significant disparities in meteorological sensitivity across aggregators with nonlinear relationships dominating the factor-load interactions. This study fills the research gap in meteorology-sensitive sub-sector load forecasting and provides theoretical / technical support for sector-level grid scheduling and building–energy system co-optimization.

Fields of Study & Subjects

Energy Load and Power Forecasting Smart Grid Energy Management Solar Radiation and Photovoltaics

Publication details

Journal / source
Journal of Physics Conference Series
Volume
3192
Issue
1
Pages
012003–012003
Publication date
1 March 2026
Language
EN