Optimizing the Pressurized Irrigation Networks Using Fuzzy Linear and Dynamic Programming

EXTENDED ABSTRACT The aim of this study was the optimization of pressurized irrigation networks using conventional and fuzzy numbers. In particular, the method of dynamic programming was utilized, whose implementation required the development of a program in Visual Fortran. The program was first applied using conventional numbers in three irrigation networks in the area of Kavasila (the former Lake Giannitsa) and the results were compared with those obtained by linear programming. Fuzzy logic was then applied to the same problem, since some parameters of the problem contain uncertainty. The problem was first approached utilizing fuzzy linear programming and consequently a fuzzy dynamic programming methodology was developed for the case of irrigation networks. The parameter with uncertainty in our problem was assumed to be the pressure at hydrants. The conclusions were as follows: a) in case of solving irrigation networks by conventional numbers, the method of dynamic programming gives the best results for irrigation networks having more than 80 pipelines, b) for smaller networks, the use of linear programming is preferable, c) in case of fuzzy numbers, the fuzzy dynamic programming is a very efficient method, fully satisfying the uncertainties of the problem and presenting a good flexibility with respect to the size of calculations, especially for networks having more than 40 pipelines.