Key points are not available for this paper at this time.
The forecasting of flooding plays a pivotal role in the field of hydrology and serves as an essential measure for preventing any possible flood damage. This study presents an analysis of the utilization of an Adaptive Neuro-Fuzzy Inference System (ANFIS) to prevent floods in river basins. ANFIS represents a hybrid approach that combines elements of neural networks and fuzzy logic methodologies for the purpose of analyzing and comparing input and output data. Performance evaluation of ANFIS is based on the ratio of Discrepancy Ratio (D) along with the determination coefficient (R2), Coefficient of correlation (R), and Root Mean Square Error (RMSE). Results: The best ANFIS model was the one that gave the best results during training. The RMSE was 622.43, with a correlation coefficient of 0.93 and an R2 of 0.86. Nevertheless, its validation performance demonstrated slightly elevated RMSE figures and lower correlation and R2 values, with RMSE at 2847.99, R at 0.91, and R2 at 0.83. An Artificial Neural Network (ANN) model with three transfer functions (SIGMOID, HYPERBOLIC TANGENT, and LINEAR) is developed and evaluated based on R = correlation and MSE = mean squared error. Furthermore, the study also entails the development of a flood model for the research area, consisting of 38 cross profiles for a detailed analysis of the river’s profile and flood inundation. This analysis aims to provide precise recommendations for flood protection measures.
Patel et al. (Fri,) studied this question.