Evaluation of Electric Vehicle Sales Forecast from Sustainability Perspective Using Time Series Analysis


Merdin D., Alkan Y. K., KABAK M.

7th International Conference on Intelligent and Fuzzy Systems, INFUS 2025, İstanbul, Türkiye, 29 - 31 Temmuz 2025, cilt.1530 LNNS, ss.30-36, (Tam Metin Bildiri) identifier

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Cilt numarası: 1530 LNNS
  • Doi Numarası: 10.1007/978-3-031-98565-2_4
  • Basıldığı Şehir: İstanbul
  • Basıldığı Ülke: Türkiye
  • Sayfa Sayıları: ss.30-36
  • Anahtar Kelimeler: Deep Learning, DeepAR Model, Electric Vehicle, Time Series Analysis
  • Gazi Üniversitesi Adresli: Evet

Özet

Electric vehicles have become one of the preferred means of transportation as they contribute to the fight against environmental pollution by reducing the use of fossil fuels. Electric vehicle sales have a dynamic structure that is constantly changing due to reasons such as public support, government commitments to reduce carbon footprint, buyer behavior, etc. To make better sales, forecasting is critical for the correct direction of investments, material supply, personnel employment, and sustainability. Time series analysis is a common method used in decision-making processes that examines variables’ causality and dynamic relationships. Due to the better sales forecasting of electric vehicles, multiple time series analysis and probabilistic analysis. For this purpose, the DeepAR method, developed by Amazon, was used to analyze multiyear electric vehicle sales data for countries. With the results obtained from this study, potential electricity consumption can be evaluated, and the carbon emissions of vehicles can be determined. In addition, the study is expected to contribute to researchers in determining the sales volume of electric vehicles by country and the number and location of the required charging stations.