Akademik Veri Yönetim Sistemi
Academic Platform journal of engineering and smart systems (Online), cilt.13, sa.1, ss.1-6, 2025 (TRDizin)
This study analyzes the detection of security attacks on smart vehicles using the Exponentially Weighted Moving Average (EWMA) algorithm. We employed synthetically generated datasets, consisting of 80% non-attack and 20% attack scenarios. Various smoothing parameters (α\alphaα) were tested within the EWMA framework, specifically at values of 0.8, 0.7, and 0.6, with 0.7 yielding the most promising results. In our analysis, we normalized the selection function in the EWMA algorithm based on expert evaluations to establish the impact of different factors on anomaly detection. Specifically, we assigned weights of 24% to RPM, 40% to speed, and 18% each to fuel quantity and accelerator pedal position. The results demonstrate that the EWMA algorithm can effectively issue warnings for vehicles under potential attack, enabling proactive measures to mitigate security risks. This research contributes to enhancing the safety and reliability of smart vehicles by facilitating timely responses to detected security threats.