Neuromodeling in Irrigation Management for Sustainable Agriculture

Abstract

The article presents the results of research on possibility and efficiency of introducing neuromodeling in irrigation control systems. One of the most effective methods of reducing water supply, saving irrigation water and, as a consequence, the sustainability of agriculture, is the use of differentiated irrigation cycles. However, traditional approaches based only on the physical modeling of processes and relationships, on the one hand, often make it difficult to find effective solutions, and on the other, are difficult to put irrigation into practice. New data mining tools provide improved accuracy and simplicity of implementation by resolving complex relationships in large amounts of parameters and have great potential. In this regard, it seems appropriate to use methods of neural network data analysis. An approach based on a data mining model is proposed, namely a Kohonen neural network clustering model and GIS technologies.