Hybrid CNN-LSTM Model for Multi-industry Electricity Demand Prediction

Abstract

Accurately predicting electricity demand is crucial for optimizing power resource allocation, improving the safety and economic performance of power grid operations, and providing significant economic and social benefits. To address this challenge, we propose a hybrid model that combines Convolutional Neural Networks (CNNs) and Long Short-Term Memory (LSTM) networks for predicting electricity demand across different industries in urban areas. The proposed model leverages the LSTM component to capture the temporal patterns of the time series data and the CNN component to extract spatial features of electricity demand across different areas. We evaluate our model on a diverse dataset of electricity demand from multiple city areas and industries. The experimental results demonstrate that our proposed model outperforms state-of-the-art methods, resulting in significant improvements in the accuracy of electricity demand prediction. Overall, our proposed hybrid model provides a valuable framework for accurately predicting electricity demand and has practical implications for power grid operations and management in urban areas.