Zastosowanie sztucznych sieci neuronowych do krótkoterminowej prognozy przepływów wezbraniowych w rzekach

• Autorzy: Siuta T.

Warianty tytułu

EN

Artificial neural network application in short-term river freshet flows forecasting

• Języki publikacji: PL

Abstrakty

EN

Recently the effects of global warming are more and more visible. The excent of enormous floods and their magnitudes are very difficult or even impossible to foresee, basing only on the medium-weather forecast. It often occurs that the reliable scenario can be obtained only when basing on the observations of the behavior of the river's catchment regarding its response to the precipitation in the real-time. From the point of view of the developing anti-flood warning systems it is important to know not only the longest possible to foresee horizon but also the degree of improvement of effectiveness of short-term prognosis. In the paper the possibilities of the usage of the artificial neural networks to determine the time-space relations between unsteady flows in the river's catchment are discussed. The usage of the neural nets is also proposed to identification of the exact time of occurrence of the culmination flows in the outlet cross-section. The proposed method is tested on the two examples of partial catchment of the river Raba, in the case of a short term prognosis. In the paper the possibilities and frontiers of the artificial neural networks' ability to foreseeing of the behavior of the natural system are also taken under consideration, in the case of different types of natural river systems.

Słowa kluczowe

PL

Dobczyce; Kasinka Mała; rzeka Raba; sztuczne sieci neuronowe; transformacja fali wezbraniowej

• Wydawca: Wydawnictwo Politechniki Krakowskiej im. Tadeusza Kościuszki
• Czasopismo: Czasopismo Techniczne. Środowisko
• Rocznik: 2002
• Tom: R. 99, z. 2-Ś
• Strony: 103--119
• Opis fizyczny: Bibliogr. 10 poz., tab., wykr., wz., il.

Twórcy

autor

Siuta T.

Bibliografia

• [1] M. Campolo, P. Andreussi, A. Soldati, River flood forecasting with a neural network model, Water Resources Research, 35, 1999, pp. 1191-1197.
• [2] J. Hertz, A. Krogh, R.G. Palmer, Wstęp do teorii obliczeń neuronowych, WNT, 1995.
• [3] K. Hsu, H. Gupta, S. Sorooshian, Artificial neural network modelling of the rainfall-runoff process, Water Resources Research, 31, 10, 1995, pp. 2517-2530.
• [4] A.W. Jayawardena, D.A.K. Fernando, Artyficial neural networks in hydro-meteorological modelling, Developments in Neural Networks and Evolutionary Computing for Civil and Structural Engineering, 1995, pp. 115-120.
• [5] P. Kneal, L. See, Developing a Neural Network for Flood forecasting in the Northumbria Area of the North East Region, Environment Agency, Final Report, University of Leeds, 1999.
• [6] J. Moody, J. Utans, Principled architecture selection for neural networks: aplication to corporate bond rating prediction, Advances in Neural Inform. Processing Systems 4, 1992, pp. 683-690.
• [7] J. R. Searle, Minds, Brains and Programs, The Behavioral and Brain Sciences, 3, 1980.
• [8] L. See, R. Abrahart, S. Openshaw, An integrated neuro-fuzzy statistical approach to hydrological modelling, Proceedings of the Third International Conference on GeoComputation, University of Bristol, 17-19 September 1998.
• [9] Saxen Bjorn, NNDT- Neural Network Development Tool, http://www.it.uom.gv/pdp/DigitalLib/Neural/Neu.soft.htm, 1995.
• [10] R. Tadeusiewicz, Sieci neuronowe, Akademicka Oficyna Wydawnicza, Warszawa 1993.