Parthasarathi Choudhury

Dynamic ANN Modeling for Flood Forecasting in a River Network

An experiment on predicting flood flows at each of the upstream and a down stream section of a river network is presented using focused Time Lagged Recurrent Neural Network with three different memories like TDNN memory, Gamma memory and Laguarre memory. This paper focuses on application of memory to the input layer of a TLRN in developing flood forecasting models for multiple sections in a river system. The study shows the Gamma memory has better applicability followed by TDNN and Laguarre memory.

Flood Forecasting in River System Using ANFIS

The aim of the present study is to investigate applicability of artificial intelligence techniques such as ANFIS (Adaptive Neuro‐Fuzzy Inference System) in forecasting flood flow in a river system. The proposed technique combines the learning ability of neural network with the transparent linguistic representation of fuzzy system. The technique is applied to forecast discharge at a downstream station using flow information at various upstream stations. A total of three years data has been selected for the implementation of this model. ANFIS models with various input structures and membership functions are constructed, trained and tested to evaluate efficiency of the models. Statistical indices such as Root Mean Square Error (RMSE), Correlation Coefficient (CORR) and Coefficient of Efficiency (CE) are used to evaluate performance of the ANFIS models in forecasting river flood. The values of the indices show that ANFIS model can accurately and reliably be used to forecast flood in a river system.

PGP-MIP model for evaluating non-damaging flows in a river system

A new pre-emptive goal programming model formulation accounting sequential and gradual changes in flow depths/discharge at a river section is presented. The model is used for evaluating conditions for maximum safe drainage for a river basin in India. The model is applied to Barak river system in India for determining upstream flow sets satisfying downstream flood flow criterion. To account for nonlinearity in flood damage function downstream flow rates are partitioned into a number of flow zones. Logical constraint equations are used to ensure gradual and sequential movement of water in the flow zones. The model is used to determine safe drainage capacity for giving higher priority to basin drainage and downstream flood safety respectively. As determined by the river system properties, the study results reveal the upstream flow combinations satisfying downstream safe flow criterion; depicts nonlinear variations of the sub basins flow with the peak flow rate at the main channel. The study demonstrates applicability of the PGP-MIP model with results that are useful to the study river system.

Simulation and forecasting of downstream flow top width in a river reach using upstream flows

In the present study, a modified Muskingum model incorporating flow and flow top width variables for a river reach and a system of reaches has been developed. The model assumes power form relationship between water discharge and flow top width at a section. The model is used to simulate and forecast flow top width on the basis of upstream flow(s) in Barak basin in India. Flow top width corresponding to different flow depths at the downstream section is estimated using DEM. Parameters in the proposed models are estimated applying NSGA-II. Model results are evaluated with standard statistical criteria. Application of the models show that for the estimated parameters, modified Muskingum models result to predicted downstream flow top width values that closely follow with the observed data. Results also indicate that the modified multiple inflows Muskingum model is more efficient and can be used with multiple upstream flows to simulate and forecast downstream flow top width in a river system.

Application of multi‐objective technique in modeling water and sediment flow in river reaches

Usually water resources problems consist of multiple objectives that may be conflicting and competing in nature. To evaluate optimal water resources system performances often it is required to obtain a compromise solution satisfying several goals and objectives. For example, in the case of multipurpose reservoir operations, a number of conflicting and competing purposes such as supply of water for conservation uses, downstream flood control, hydropower generation and related environmental objectives are to be optimally satisfied. It may be noted that for deriving maximum benefit from conservation uses reservoir storage should be as high as possible; on the other hand to achieve maximum flood control benefits the storage should be kept as low as possible. Since flood control and conservation objectives are conflicting in nature, higher achievement in flood control objective results in lower attainment of the conservation objectives. In other areas of water resources such as, rainfall runoff modeling, water...