ANN-Assisted Planning of Conjunctive Use of Canal and Ground Water in Canal Commands
Abstract
Presented herein is an algorithm for ANN‐assisted planning of the optimal cropping pattern and the associated groundwater development in a canal command area, followed by an illustration. The planning in the illustration ensures the maximization of the cropped area subject to the constraint of limiting the maximum water table depth to an acceptable limit. A multilayer feed forward ANN model, relating the maximum water table depth to the crop areas, is trained using back‐propagation learning algorithm. The data invoked for the training comprise an array of the cropping patterns, and the corresponding maximum water table depths. These data are generated through a pre‐calibrated numerical model of groundwater flow. Subsequently, the trained ANN model is linked to a GA based optimizer for arriving at the optimal cropping pattern and the associated pumping pattern.
Authors: Deepak Kashyap, Susmita Ghosh, Jane W. Z. Lu, Andrew Y. T. Leung, Vai Pan Iu, Kai Meng Mok
Published in: AIP conference proceedings (2010)