Open AccessResearch on Forecasting Model of Daily Discharge in Karst Area Based on Mea Grey Neural Network
Author(s) -
Xia Li,
Xin Jin,
Peng Guo
Publication year - 2020
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1549/2/022087
Subject(s) - artificial neural network , surface runoff , karst , laminar flow , maxima and minima , flow (mathematics) , computer science , nonlinear system , turbulence , watershed , environmental science , hydrology (agriculture) , geology , artificial intelligence , meteorology , machine learning , mathematics , engineering , geotechnical engineering , geography , ecology , paleontology , mathematical analysis , physics , geometry , quantum mechanics , aerospace engineering , biology
The water-bearing system in the karst area is complex and changeable. Water in the water-bearing medium has the characteristics of coexistence of fissure flow and pipeline flow, coexistence of laminar and turbulent flow, coexistence of linear and nonlinear flow, coexistence of continuous flow and isolated water body. In areas where economic development is relatively backward, most areas lack data or no data. Based on the characteristics of the karst area in the southwest, this paper proposes a thinking evolution algorithm to optimize the gray neural network model. This method improves the optimization ability of runoff prediction models and effectively overcomes human neural network learning. The speed is slow and there are inherent shortcomings of local minima. After studying the daily flow forecast of Zhenlong Station, it is shown that the relative error of the prediction is small and can be effectively used for short-term runoff prediction.