view:17191 Last Update: 2019-10-22
Mehran Rostamloo, Hasan Ojaghloo and Masoud Karbasi
Compartion performance of adaptive neuro-fuzzy inference system (ANFIS) and gene expression programming (GEP) to estimate water distribution uniformity coefficient in classic sprinkle irrigation systems.
One of the strategies for efficient use of water resources in the country is development of new irrigation systems including pressurized irrigation methods. Field evaluation and precise knowledge of pressurized irrigation systems performance in different terms plan as an important tool in the correct development of these systems are considered. One of the most important performance evaluation criteria in the design of pressurized irrigation systems and especially classic irrigation systems is water distribution uniformity index. On the other hand, field measurements of water distribution uniformity index in different climatic conditions, hydraulic and projects executive specifications requires spending time and too much costs and so use of indirect methods such as intelligent models can be useful. Checking the studies in the field of water distribution uniformity coefficient simulation in sprinkle irrigation systems showed no study in the field of estimate the water distribution uniformity coefficient in sprinkle irrigation by using of adaptive neuro- fuzzy inference system (ANFIS) method has been done. Therefore, present study aimed to check the performance of adaptive neuro- fuzzy inference system and comparison with results of gene expression programming method to estimate the water distribution uniformity coefficient in solid classic sprinkle irrigation system in different climatic conditions, hydraulic and physical of classic sprinkle irrigation systems was performed. Research doing method was consisting of two parts; field measurement and simulations by using of anfis and gene expression programming intelligent models. For this purpose, a solid classic sprinkle irrigation system with considering of different arrangements of pipes and sprinklers were designed and performed. Then 54 field experiment to evaluate the performance of a solid classic sprinkle irrigation system were performed. In each experiment to determine the water distribution uniformity coefficient, from a network of sampling water cans were used. Input parameters of adaptive neuro- fuzzy inference system and gene expression programming were combination of climatic factors (including; average of temperature, relative humidity, average of wind speed and direction), physical factors including; arrangement and different distances of sprinklers and sprinkler model (amount of output volumetric flow rate). Output parameter in all simulations, were water distribution uniformity coefficient (cu) in percent. 70 percent of field data obtained for learning of models and 30 percent of data for the test of models were used. In order to compare and evaluate the performance of intelligent models in estimate the water distribution uniformity coefficient of classic sprinkle irrigation systems, were used from pearson correlation coefficient statistics, root mean square error and mean absolute error. The results showed that, generally with increasing laterals distance and wind speed, amount of water distribution uniformity coefficient decreases. Field observation were indicated in the same terms, Ambo’s model sprinkler had more suitable performance of water distribution uniformity than vyr-155. The simulation results showed that the use of all input data includes; volumetric flow rate of sprinkler, sprinklers distances, wind speed, wind direction, relative humidity and average of temperature and as well as considering the effective radius equal one, lead to the best results. So that in ANFIS model, maximum amount of correlation coefficient (R) and root mean square error (RMSE) for the test phase were obtained equal to 0.77 and 7.7 %, respectively. By eliminating the parameters of relative humidity and average of temperature from model inputs, were not observed significant change in the results while by eliminating wind speed parameter, results of model's output, including; RMSE index was changed significantly. This matter showed that, water distribution uniformity coefficient values in the experiments conducted is strongly influenced by wind speed. Best performance of gene expression programming model also was related to combine the input data includes; volumetric flow rate of sprinkler, sprinklers distance, wind speed and direction, so that maximum amount of correlation coefficient in the test phase equal to 0.72 and the lowest amount of RMSE equal to 7.13% were achieved. One of the advantages of gene expression programming model than ANFIS model and other intelligent models, is offering the optimal mathematical equation between the dependent variable of uniformity coefficient and other independent variables (input variables). Generally, performance difference was little between two methods of gene expression programming and adaptive neuro- fuzzy inference and sensitivity of models showed, the temperatures and wind speed had the lowest and the most effect on water distribution uniformity coefficient changes, respectively. Estimated amounts of checking the water distribution uniformity coefficient showed that intelligent models as well as factors effect such as wind speed and sprinklers distances have been able to simulate the reducing amount of water distribution uniformity.