Atıf Sayısı
3
Görüntüleme
30
İndirme
3
Su kaynaklarının sürdürülebilir olması için periyodik nehir akım ölçümlerinin yapılması gerekmektedir. Bunun için farklı tahmin yöntemlerine gereksinim duyulmaktadır. Bu çalışmada, Yapay Zeka yöntemlerinden Derin Öğrenme (DL) ile Aksu Nehri akımlarının LSTM (Uzun-Kısa Süreli Bellek) sinir ağı ile tahmini yapılmıştır. Çalışmada, Aksu Nehri üzerindeki D20A002 No’lu Başpınar Akım Gözlem İstasyonuna (AGİ) ait 2000-2019 yılları arasını kapsayan veriler analiz için girdi olarak kullanılmıştır. Ayrıca, Tekil Spektrum Analizi’nin (TSA) LSTM’ye olan perfonmans etkisi irdelenmiştir. TSA-LSTM modeline iyileştirici olarak Adam, Adamax ve AdaGrad algoritmaları uygulanmıştır. Tahmin ve gerçek değerler karşılaştırılarak en doğru tahmin modeli belirlenmiştir. Akım tahmininde iyi performansı Adamax iyileştiricisinin sağladığı görülmüştür. TSA-LSTM modeli katsayısı (R2) tayini test aşamasında 0,9851 bulunmuştur. Elde edilen sonuçlar incelendiğinde TSA-LSTM modelinin akım çalışmaları tahmininde daha iyi sonuçlar verdiği görülmüştür.
Periodic river flow measurements must be carried out in order for water sources to be sustainable. This requires different methods of prediction. In this study, the methods of artificial intelligence were predicted by the LSTM (Long-Long-Term Memory) nerve network of deep learning (DL) and the Axu River streams. In the study, the data covering the 2000-2019 years of the D20A002 Headquarters Stream Observatory (GI) on the Aksu River was used as a input for analysis. Furthermore, the perfonmans effect of Single Spectrum Analysis (TSA) on LSTM has been exhausted. The TSA-LSTM model has been healed by Adam, Adamax and AdaGrad algorithms. The most accurate prediction model is determined by comparing the forecast and real values. It has been shown that good performance in current forecast is provided by the Adamax healer. The TSA-LSTM model ratio (R2) was found at 0.9851 in the test stage. When the results were studied, the TSA-LSTM model was shown to give better results in the forecast of current work.
Periodic river flow measurements are required to ensure sustainable water resources. For this, different estimation methods are required. In this study, Deep Learning (DL) and Aksu River flows were estimated by LSTM (Long-Short Term Memory) neural network, which is one of the Artificial Intelligence methods. In the study, the data belonging to Başpınar Flow Measurement Station (FMS) (D20A002) on Aksu River between 2000-2019 were used as input for analysis. In addition, the performance effect of Single Spectrum Analysis (TSA) on LSTM was examined. Adam, Adamax and AdaGrad algorithms were applied to the TSA-LSTM model. The most accurate estimation model has been determined by comparing the estimate and actual values. It has been observed that the Adamax optimizer provides the best performance in flow estimation. TSA-LSTM model coefficient (R2) determination was found to be 0.9851 during the test phase. When the obtained results were examined, it was seen that the TSA-LSTM model gave better results in estimating flow studies.
Alan : Fen Bilimleri ve Matematik; Mühendislik
Dergi Türü : Uluslararası
Makale Detay 
Atıf Yapanlar
Benzer Makaleler
PDF Görüntüle
Dergi Bilgisi
Eseri Dinleyin
Bu Sayfayı Yazdırın
Paylaş
Mendeley
