Landing of Autonomous UAVs on Maritime Platforms Using LSTM Layered Neural Network-Based Motion Predictor and Energy-Efficient Landing Trajectory Optimization LSTM Katmanli Sinir A?gi Tabanli Hareket Tahmincisi ve Enerji Verimli ?Iniş Yörüngesi Optimizasyonu Kullanarak Otonom ?IHA'larin Deniz Platformlarina ?Inişi
34th Signal Processing and Communications Applications Conference, SIU 2026, İstanbul, Türkiye, 7 - 10 Temmuz 2026, (Tam Metin Bildiri)
- Yayın Türü: Bildiri / Tam Metin Bildiri
- Doi Numarası: 10.1109/siu71813.2026.11636449
- Basıldığı Şehir: İstanbul
- Basıldığı Ülke: Türkiye
- Anahtar Kelimeler: energy-optimal control, landing optimization, LSTM, ship motion prediction, STFT
- Yıldız Teknik Üniversitesi Adresli: Evet
Özet
This study analyzes the short-term motion prediction of a ship platform operating under irregular sea wave conditions and evaluates the energy efficiency of landing algorithms for a quadrotor unmanned aerial vehicle. A synthetic dataset is generated using a response amplitude operator based irregular wave simulation. Spectral features obtained via the Short-Time Fourier Transform (STFT) are incorporated into the artificial neural network feature vector, and the use of a one-frame STFT is shown to significantly improve position prediction accuracy. In addition, energy-optimal and two-stage landing algorithms are compared based on a quadratic cost function. The results demonstrate that the energy-optimal algorithm is more efficient in terms of actual energy consumption than the two-stage approach.