According to the United Nations Environment Programme, the COVID-19 pandemic has created challenges for the economy and the energy sector, as well as uncertainty for the renewable energy industry. However, the impact on renewable energy during the pandemic has not been consistently determined. Instead of relying on data from year-to-year comparisons, this study redesigned the analytical framework for assessing the impact of a pandemic on renewable energy. First, this research designed an "initial prediction-parameter training-error correction-assignment combination" forecasting approach to simulate renewable energy consumption in a "no pandemic" scenario. Second, this study calculates the difference between the "pandemic" and "no pandemic" scenarios for renewable energy consumption. This difference represents the change in renewable energy due to the COVID-19 pandemic. Various techniques such as nonlinear grey, artificial neural network and IOWGA operator were incorporated. The MAPEs were controlled to within 5% in 80% of the country samples. The conclusions indicated that renewable energy in China and India declined by 8.57 mtoe and 3.19 mtoe during COVID-19 period. In contrast, the rise in renewable energy in the US is overestimated by 8.01 mtoe. Overall, previous statistics based on year-to-year comparisons have led to optimistic estimates of renewable energy development during the pandemic. This study sheds light on the need for proactive policy measures in the future to counter the global low tide of renewable energy amid COVID-19.
Keywords: COVID-19; Neural network model; Renewable energy; Scenario estimation.
© 2022 Elsevier Ltd. All rights reserved.