Energy Consumption Forecasting in Commercial Buildings during the COVID-19 Pandemic: A Multivariate Multilayered Long-Short Term Memory Time-Series Model with Knowledge Injection

The COVID-19 pandemic and the subsequent implementation of lockdown measures have significantly impacted global electricity consumption, necessitating accurate energy consumption forecasts for optimal energy generation and distribution during a pandemic. In this paper, we propose a new forecasting model called the multivariate multilayered long short-term memory (LSTM) with COVID-19 case injection ( mv − M − LSTM − CI ) for improved energy forecast during the next occurrence of a similar pandemic. We utilized data from commercial buildings in Melbourne, Australia, during the COVID-19 pandemic to predict energy consumption and evaluate the model’s performance against commonly used methods such as LSTM, bidirectional LSTM, linear regression, support vector machine, and multilayered LSTM (M-LSTM). The proposed forecasting model was analyzed using the following metrics: mean percent absolute error (MPAE), normalized root mean square error (NRMSE), and R 2 score values. The model mv − M − LSTM − CI demonstrated superior performance, achieving the lowest mean percentage absolute error values of 0.061, 0.093, and 0.158 for DatasetS 1 , DatasetS 2 , and DatasetS 3 , respectively. Our results highlight the improved precision and accuracy of the model, providing valuable information for energy management and decision making during the challenges posed by the occurrence of a pandemic like COVID-19 in the future.