Bulut, MerveAydilek, HuseyinErten, Mustafa YasinOzcan, Evrencan2025-01-212025-01-2120250952-19761873-6769https://doi.org/10.1016/j.engappai.2024.109602https://hdl.handle.net/20.500.12587/24138Climate change across the globe, especially extreme temperature events, is increasing pressures on energy systems. The extraordinary situation that California and the West faced in late August and early September 2022, when record temperatures led to a spike in electricity demand, provided an important backdrop for the resilience and sustainability of clean energy technologies. The electricity market managed by the California Independent System Operator is considered in this study to examine the potential impacts on electricity demand spikes and system resilience. The methodology of the research involves analyzing the system operator's responses to electricity demand using advanced deep learning algorithms, convolutional neural network - long-short term memory and attention mechanism models. 1, 3 and 7-days forecasts of electricity demand were made using models in the day ahead market. In 1-day forecasts, while the former models have a mean absolute percentage error value of 12.40%, the latter model has a lower error rate of 10.36%. Overall findings obtained from various scenarios show that the long-short term memory - attention mechanism can more effectively understand complex patterns in energy demand and has the potential to increase system stability against such extreme weather events. The advanced horizon of the study offers an important perspective on how clean energy technologies, especially battery energy storage systems, can provide solutions to today's priority problems such as climate change and extreme temperature.eninfo:eu-repo/semantics/closedAccessDay-ahead market; Battery energy storage systems; California independent system operator; Forecast models; Attention mechanismArticle13910.1016/j.engappai.2024.1096022-s2.0-85209243677Q1WOS:001361117700001N/A