Open access

Comparative Study of LSTM, GRU, and CNN-Based Hybrid Models for Short-Term Load Forecasting

Muath E. Alharbi Sreerama Kumar A. Milyani

Jul 2026 · European Journal of Engineering and Technology Research · Vol 11, pp. 27-33 · 0 citations · 12 references

TL;DR

The results indicate that GRU consistently achieves the best performance in terms of mean absolute percentage error, whereas CNN-RNN hybrid models show sensitivity to the depth of architecture and choice of activation function.

Abstract

This study investigates the effectiveness of long short-term memory (LSTM) networks, Gated Recurrent Unit Networks (GRU), and their hybrid models with a Convolutional Neural Network (CNN), namely CNN-GRU and CNN-LSTM, in the short-term electrical load forecasting of practical power systems. This study uses hourly load data spanning 11 years from 2014 to 2024 from a typical regional transmission network. A set of experiments was conducted using various input configurations, model complexities, and activation functions. The models were tested in univariate and multivariate settings, including ten input features and recurrent layer variations. The results indicate that GRU consistently achieves the best performance in terms of mean absolute percentage error, whereas CNN-RNN hybrid models show sensitivity to the depth of architecture and choice of activation function. These deep learning methods were compared in terms of their model complexity and forecasting accuracy to provide guidance for selecting the optimal models for short-term electrical load forecasting.

DOI Semantic Scholar

Read PDF

Similar papers

Open access Aug 2026

Comparative Analysis of LSTM and Hybrid Attention-Based Architectures for Short-Term Electricity Consumption Forecasting: A Case Study in Togo

Overall, the study demonstrates that recurrent deep learning architectures provide reliable and operationally relevant solutions for electricity consumption forecasting in highly structured energy demand series.

Séna Apeke, Yao Bokovi, K. Gbafa et al. · Science Journal of Energy En... · 0 citations

Aug 2026

Development of a Hybrid CNN-BILSTM Model with Attention Mechanism for Predicting Thermal Load in District Heating Systems

Xu-Yang Cui, Yun-Zhe Zhang, Jun-Lan Yang et al. · Thermal Engineering · 0 citations

Conference Jul 2026

Spatio-Temporal Cab Demand Forecasting using Hybrid LSTM-CNN Models

The research is aimed at designing and implementing an intelligent cab demand forecasting model that utilizes the methods of Long Short-Term Memory (LSTM) and Convolutional Neural Network (CNN) models so that future ride demand can be accurately forecasted in an urban environment. Two groups were studied in this compar...

B. Rajesh, P. Priyadharshini, K. Balaji et al. · 2026 7th International Confe... · 0 citations

Open access Aug 2026

Deep Learning for VIX Volatility Forecasting: A Comparative Study of LSTM, BiLSTM, and CNN-LSTM Architectures

The CNN-LSTM model consistently outperforms both unidirectional and bidirectional LSTMs, and consistently outperforms both unidirectional and bidirectional LSTMs during periods of market turbulence, with CNN-LSTM demonstrating the strongest resilience.

Siyi Zheng · Applied and Computational En... · 0 citations

Open access Jul 2026

Mixed non linear programming model for optimizing and training long short-term memory architectures: application in traffic flow forecasting

It is demonstrated that framing LSTM architecture optimization as a mixed-variable combinatorial problem, coupled with PSO-based optimization, substantially improves forecasting performance, offering a robust and versatile strategy for accurate traffic prediction and other complex sequential data applications.

Taoufyq Elansari, Hamza H. Sulimani · Evolutionary Systematics · 0 citations

Open access Aug 2026

Hybrid deep learning model for enhanced short-term gold price forecasting

STGP-Net is introduced, a novel hybrid deep learning model designed to enhance prediction accuracy by integrating one-dimensional convolutional neural network (1D-CNN) and long short-term memory (LSTM) that provides better performance and robustness than alternative hybrid models tested.

Hoang Ha Nguyen, Minh Duc Nguyen, Cuong H. Nguyen-Dinh · IAES International Journal o... · 0 citations