ANALYSIS OF SEDIMENT DEPOSITION OF BRAHMANI RIVER BY NEURAL NETWORK AND MULTIPLE LINEAR REGRESSION

Abstract

Estimating the settled sediment production is critical for water resource planning and management, as well as environmental protection. Sediment creation, transit, and the resulting sediment load in rivers are all influenced by environmental change. The natural sediment flow in the river is assessed in terms of sediment concentration using an artificial neural network. This is accomplished by teaching the network how to gather natural stream data from reputable sources. Choosing the right neural network structure the performance results based on two sorts of indicators, viz. correlation coefficient (R2

), root mean square error (RMSE). Internal suspicions are not explicitly built-in traditional techniques of estimating sediment production (e.g. regression models). This model, on the other hand, is unable to augment intellectual capacity of the underlying connections between the data gathered or to estimate the influence of each sediment yield component. It is easier and fewer expensive to develop artificial neural networks for sedimentation prediction and predicting factors. Environmental change has an impact on sediment production, transit, and the consequent sediment load in rivers. The Brahmani river basin was contrast using multiple linear regression and artificial neural networks in this study. This study used a four-year data collecting period and the back-propagation approach. The finest model findings will be used as an estimate in future hydrological structural blueprint studies.