Forecasting Of Stream Flow Usıng Singular Spectrum Analysis And Long-Short Term Memory Networks

Abstract

Stream flow estimation has an important role in the planning and management of water resources. Accurate estimation of stream flow data, that is characterised ny non-linear and non-stationary, is a challenging problem. In recent years, data-based techniques have been used extensively in forecasting of stream flow. In this study, stream flow estimation was made with the Long-Short Term Memory (LSTM) Networks from Deep Neural Networks, which were used as popular. Subband data was obtained by using Single Spectrum Analysis (SSA), which plays an important role in the analysis of time series in order to increase the forecast performance. As a result of estimation of SSA subband data of stream flow forecasting data with LSTM network, one ahead forecasting study was carried out. Using this proposed SSA-LSTM model, high performance forecasted data was obtained with 0.0021 Mean Square Error (MSE) value, 0.0361 Mean Absolute Error (MAE) value and 0.9710 Correlation (R) coefficient value.

Keywords

• Stream flow forecasting,Deep Neural Networks,Long-Short Term Neural Networks,Singular Spectrum Analysis

Tekil Spektrum Analizi ve Uzun-Kısa Süreli Bellek Ağları ile Nehir Akım Tahmini

Öz

Su yapılarının planlanması ve yönetiminde nehir akım tahminleri önemli bir yere sahiptir. Lineer olmayan ve durağan olmayan karaktere sahip nehir akım verilerinin doğru tahmini zorlu bir problemdir. Son yıllarda veri tabanlı teknikler, nehir akım problemlerinde yoğun olarak kullanılmaktadır. Önerilen çalışmada popüler olarak kullanılmaya başlanan Derin Sinir Ağlarından Uzun – Kısa Süreli Bellek (Long-Short Term Memory, LSTM) Ağları ile nehir akım tahmini gerçekleştirilmiştir. Tahmin performansını artırmak üzere zaman serilerinin analizinde önemli bir yer tutan Tekil Spektrum Analizi (TSA) kullanılarak alt bant verileri elde edilmiştir. Nehir akım tahmin verisine ait TSA alt bant verilerinin LSTM ağları ile tahmini sonucu bir ileri adım tahmin çalışması gerçekleştirilmiştir. Önerilen TSA-LSTM modeli kullanılarak 0.0021 Ortalama Karesel Hata (MSE) değeri, 0.0361 Ortalama Mutlak Hata (MAE) değeri ve 0.9710 Korelasyon (R) değeri ile yüksek performanslı tahmin verisi elde edilmiştir.

Anahtar Kelimeler

• Nehir akım tahmini,Derin Sinir Ağları,Kısa-Uzun Bellek Ağları,Tekil Spektrum Analizi

Kaynakça

1. Bayazıt M, (1998). Hidrolojik Modeller, İTÜ İnşaat Fakültesi Matbaası, İstanbul.
2. Cuddington K., M. J. Forth, L. R. Gerber et al., “Process-based models are required to manage ecological systems in a changing world,” Ecosphere, vol. 4, no. 2, pp. 1–12, 2013
3. Liu, Z., Zhou, P., Chen, X., Guan, Y., 2015. A multivariate conditional model for streamflow prediction and spatial precipitation refinement. J. Geophys. Res. Atmos. 120 (19).
4. Ekmekçi M. “Hacettepe Ders Notları” ,http://www.dsi.gov.tr/faaliyetler/turkiye-ulusal-hidroloji-komisyonu (Erişim tarihi, 2017)
5. Box, G. E. P., Jenkins, G. M., 1970, Time Series Analysis: Forecasting and Control. Holden-Day, San Francisco.
6. Carlson, R. F., MacCormick, A. J. A., Watts, D. G., 1970. “Application of linear models to four annual streamflow series”, Water Resour. Res. 6 (4), 1070–1078.
7. Zhang, H., Singh, V.P., Wang, B., Yu, Y., 2016. CEREF: A hybrid data-driven model for forecasting annual streamflow from a socio-hydrological system. J. Hydrol. 540, 246–256.
8. Mehr A.D., 2018. “An improved gene expression programming model for streamflow forecasting in intermittent streams”, Journal of Hydrology, 563, 669-678.

9. Dehghani M., Seifi A., Madvar H.R., 2019 “Novel forecasting models for immediate-short-term to long-term influent flow prediction by combining ANFIS and grey wolf optimization”, Journal of Hydrology, 576, 698-725.
10. Sahoo A., Samantaray S., Ghose D. K., 2019. “Stream Flow Forecasting in Mahanadi River Basin using Artificial Neural Networks”, Procedia Computer Science, 157, 168-174.
11. He Y., Yan Y., Wang X., Wang C., 2019. “Uncertainty Forecasting for Streamflow based on Support Vector Regression Method with Fuzzy Information Granulation”, Energy Procedia, 158, 6189-6194.
12. Yu X., Wang Y., Wu L., Chen G., Wang L., Qind H., 2019. “Comparison of support vector regression and extreme gradient boosting for decomposition-based data-driven 10-day streamflow forecasting”, Journal of Hydrology, in press, 124293
13. Hadi S. J., Tombul M., 2018. “Monthly streamflow forecasting using continuous wavelet and multi-gene genetic programming combination”, Journal of Hydrology, 561, 674-687.
14. Latifoğlu L., Kisi O., Latifoglu F., 2015. “Importance of hybrid models for forecasting of hydrological variable”, Neural Computing & Applications, 26, 1669-1680.
15. Zealand CM., Burn DH., Simonovic SP., 1999, Short term streamflow forecasting using artificial neural networks, Journal of Hydrology, 214, 32–48
16. Ni L., Wang D., Singh V. P., Wu J., Wang Y., Tao Y., Zhang J., 2019. “Streamflow and rainfall forecasting by two long short-term memory-based models”, Journal of Hydrology, in press, 124296.
17. Mouatadid S., Adamowski J. F., Tiwari M. K., Quilty J. M., 2019. “Coupling the maximum overlap discrete wavelet transform and long short-term memory networks for irrigation flow forecasting”, Agricultural Water Management, 219, 72-85.
18. Dehghani M., Seifi A., Madvar H.R., 2019 “Novel forecasting models for immediate-short-term to long-term influent flow prediction by combining ANFIS and grey wolf optimization”, Journal of Hydrology, 576, 698-725.
19. LeCun, Y., Bengio, Y., Hinton, G., 2015. Deep learning. nature 521 (7553), 436
20. Ni L., Wang D., Singh V. P., Wu J., Wang Y., Tao Y., Zhang J., 2019. “Streamflow and rainfall forecasting by two long short-term memory-based models”, Journal of Hydrology, in press, 124296.
21. Latifoğlu L., Kisi O., Latifoglu F., 2015. “Importance of hybrid models for forecasting of hydrological variable”, Neural Computing & Applications, 26, 1669-1680.
22. Kisi O., Latifoğlu L., Latifoglu F., 2014. “Investigation of Empirical Mode Decomposition in Forecasting of Hydrological Time Series”, Water Resources Management, 28, 4045-4057.
23. Nourani V., Andalib G., Sadikoglu F., 2017. “Multi-station streamflow forecasting using wavelet denoising and artificial intelligence models”, Procedia Computer Science, 120, 617-624.
24. Fang W., Huang S., Ren K., Huang Q., Huang G., Cheng G., Li K., 2019. “Examining the applicability of different sampling techniques in the development of decomposition-based streamflow forecasting models”, Journal of Hydrology, 568, 534-550.
25. http://www.dsi.gov.tr/faaliyetler/akim-gozlem-yilliklari (Erişim Tarihi: Aralık 2019).
26. Golyandina N., Zhigljavsky A., 2013, Singular Spectrum Analysis for Time Series, Springer.
27. Sepp Hochreiter, Jurgen Schmidhuber, LONG SHORT-TERM MEMORY, Neural Computation, 9(8):1735-1780, 1997
28. Olah, C. (2015, 27 Mayıs), Understanding LSTM Networks, Erişim Adresi: http://colah.github.io/posts/2015-08-Understanding-LSTMs/