Optimum Layout Selection of Sprinklers Using GA

IntroductionAccording to the urban green space development, water resources reduction, and other various factors, it seems necessary to use of sprinkler irrigation in irrigating the urban green space. There is no clear and principled method to determine the sprinkler system layout of green spaces. Sprinkler layout usually has a triangular or square pattern and is estimated automatically. In this paper, using genetic algorithm, the location of Sprinkler system installation is determined. In order to provide the maximum distribution uniformity.Materials and methodsAt first, the spray pattern of Hunter A17 sprinkler on the farm of Islamic Azad University of Ferdows in Iran and the regional environmental conditions were determined, and then the sprinklers location was obtained by genetic algorithm. In this algorithm, the points on the map were elected as index points (sample containers). These points have a one-meter distance from each direction, horizontally and vertically. Genetic Algorithm is a searching method of optimization that has been inspired by human body. Chromosome structure that has been presented in the algorithm is a set of two-dimensional coordinate system of the of sprinkler location on the map. These coordinates are a binary number with eight bits, so chromosome length is 144 bits. The structure contains nine sprinklers that have nine two-dimensional coordinate systems. This plan aims is to achieve high distribution uniformity coefficient for the farm. Therefore, the objective function is a part of distribution uniformity formulation of the farm. Obtained uniformity coefficient depends on the number and spray pattern of a single sprinkler (which is depends on climatic conditions). Considering the spraying radius of sprinklers and preventing he water outgo from the field, coordinate selection is not limited to length and width of the field. Minimum distance of coordinate sprinklers location from the field border must be equal to spray radius. For this reason, border strip has supposed equal to 3, 4 or 5 meter and algorithm run with these strips. This problem was supplied by increase the distribution uniformity. The increase of the distribution uniformity will decrease deep percolation.Results and Discussion The distribution uniformity changes with variation range of border strips. If the range of border strips decreases, the distribution uniformity will increase. Considering the experimental farm, uniformity coefficient obtains 78.9 percent from genetic algorithm method, versus the coefficient for the manual method of sprinkler’s layout obtained 73.3 percent. Conclusion Genetic Algorithm is used less in irrigation science and this research tries to create a new vision to this aim. This research concluded that, Genetic Algorithm is one of the best methods in optimizing this complicated problem.