285 / 2024-02-28 23:48:42

Deriving flood routing network in mega-basin using data-driven artificial intelligence model based on flood routing pattern recognition and intelligent heuristic optimizer

Flood routing,Fuzzy clustering,Long short-term memory,Evolutionary algorithm,Mega-basin

Fast and accurate simulation of flood routing network in mega-basin is key to the development of water resources management policies. However, it is difficult for hydrologic and hydraulic methods based on physical mechanism to satisfy these requirements. To meet these practical requirements, this paper proposes a novel artificial intelligence method for deriving flood routing network in mega-basin. The proposed method uses the fuzzy clustering iteration method to identify multiple typical flood routing patterns; for all the samples within each flood routing pattern, the long short-term memory (LSTM) is utilized to simulate the nonlinear mapping relationship between the influence inputs and the target outputs, while the emerging intelligent heuristic optimizer is chosen to determine suitable parameters for the LSTM model. The feasibility of the proposed method is fully evaluated in the Upper Yangtze River Basin, a mega-basin in China. The simulation results demonstrate that the proposed method can yield better comprehensive results than the hydrologic method, especially in the high-discharge flood routing pattern. Compared to the hydraulic method, it does not require additional hard-to-access data and a lot of computation time. Hence, this study case confirms that intelligent heuristic optimizer and flood routing pattern recognition techniques can enhance the performance of a standalone data-driven artificial intelligence method in deriving flood routing network in mega-basin.