simulation of water distribution pattern of single sprinkler using fuzzy logic

abstract a model based on fuzzy logic has been developed to simulate the distribution pattern of a single sprinkler. a nelson r3000 sprayer with rotator pad and water application angle of 360o which is one of the most common types of center pivot and linear move sprinklers was selected. several field experiments according to the iso-8026 and asae-s398.1 standards were performed to assess the water distribution pattern in no-wind and windy conditions at the research center of agricultural faculty, university of tabriz-iran. results of thirty reliable experiments were used to educate & validate the model. minimum and maximum wind speeds recorded under the field conditions were 0.57 and 7.41 ms-1, respectively. in order to comparative analyze between simulated values and observations several statistical criteria like r2, cd, ef, crm, mae and rrmse were used. for the training dataset the average values of r2 and rrmse were achieved 0.98 and 0.257, respectively. also the values of parameters r2 and rrmse for the validation dataset were calculated 0.96 and 0.34 respectively. as well as the other above mentioned statistical parameters for both training dataset and validation dataset were found satisfactory. to make the results practical a model was developed as a matlab m-file, using fuzzy logic that takes wind velocity and direction as the inputs and could simulate the distribution pattern of single sprinkler and have ability to display graphical and excel file of results. with the statistical comparisons between simulated water distributions patterns with observed ones it was finally concluded that fuzzy model had excellent ability to simulate the water distribution pattern.