Please use this identifier to cite or link to this item:
|Title:||Hybrid subtractive clustering technique for estimation of ground water table|
|Abstract:||Groundwater scarcity and groundwater table decline at alarming rate create many problems for managers and planners. This problem will be worsening in future as demand for water would rise tremendously due to steady rise in human population. Optimal-exploitation of groundwater ‘in all the areas that are getting fast depleted’ is required to estimate the groundwater table. Keeping this in view, different models have been developed for an Indian catchment located in North – East region of India. Groundwater studies have not been conducted much by researchers for the problematic area of Budaun district where the water table is continuously declining at a faster rate. The present study is taken up to study the groundwater availability in the district. Models have been developed for prediction of water table behaviour and a proper groundwater recharge plan for the problematic areas can be suggested. Over the last decade, soft computing techniques like Fuzzy logic and Artificial Neural Networks (ANN) are increasingly used in hydrological studies. Furthermore, their computational speed in simulating and forecasting is very welcomed in real time operations. It is robust and flexible in managing real world complex systems involving uncertainty and imprecise data. Fuzzy Logic Controller (FLC) provides a means of converting a linguistic control strategy based on operators’ knowledge into an automatic control strategy. An important feature of fuzzy set theory is the symmetry between the objective function and constraints. During this study, a Neuro-Fuzzy approach such as ANFIS (Adaptive Neuro Fuzzy Inference System) subtractive cluster method (SCM) and ANN techniques have been used for monsoon (M) and non-monsoon (NM) periods in order to estimate the ground water tables. Input data of the network are composed by past measurements of nearby inflow and rainfall, and the quantity of water which has been pumped out from ground during that specific period. Input data are fuzzified with different degrees of membership. The models are developed and applied for Budaun district in Uttar Pradesh to get the optimum output during monsoon and non-monsoon period. Models developed by SCM technique give better results than ANN technique.|
|ISSN:||0975-2412 (Online); 0771-7706 (Print)|
|Appears in Collections:||BVAAP Vol.20(1) [June 2012]|
Items in NOPR are protected by copyright, with all rights reserved, unless otherwise indicated.