From ARIMA to Attention: Power Load Forecasting Using Temporal Deep Learning

TL;DR

This paper compares traditional and deep learning models for short-term power load forecasting, demonstrating that Transformer models outperform ARIMA, LSTM, and BiLSTM in accuracy and robustness on PJM data.

Contribution

It provides an empirical evaluation of ARIMA, LSTM, BiLSTM, and Transformer models, highlighting the superior performance of attention-based architectures in power load forecasting.

Findings

Transformer achieved 3.8% MAPE, outperforming other models.

Attention-based models better capture complex temporal patterns.

Deep learning models improve forecasting accuracy and robustness.

Abstract

Accurate short-term power load forecasting is important to effectively manage, optimize, and ensure the robustness of modern power systems. This paper performs an empirical evaluation of a traditional statistical model and deep learning approaches for predicting short-term energy load. Four models, namely ARIMA, LSTM, BiLSTM, and Transformer, were leveraged on the PJM Hourly Energy Consumption data. The data processing involved interpolation, normalization, and a sliding-window sequence method. Each model's forecasting performance was evaluated for the 24-hour horizon using MAE, RMSE, and MAPE. Of the models tested, the Transformer model, which relies on self-attention algorithms, produced the best results with 3.8 percent of MAPE, with performance above any model in both accuracy and robustness. These findings underscore the growing potential of attention-based architectures in accurately capturing complex temporal patterns in power consumption data.