بهینه سازی برنامه ریزی آبیاری در شرایط مختلف تامین آب با استفاده از الگوریتم مورچگان

The optimization of agricultural water consumption is very important due to the limitation of water resources and its value in agriculture. Two methods of field experiments and simulation models can be used to determine the best irrigation plan. The first method, in addition to requiring multiple field trials, also includes limitations such as limitation of location and conditions of the test and the short time of running test. On the other hand, simulation models are a suitable choice to improve management of water consumption in the field due to the quantitative impacts of water on crop performance. In this study, the optimization of water consumption and dominate cropping pattern for five major crops of winter wheat, barley, forage corn, grain corn and canola, in the Qazvin plain irrigation network were evaluated.
Qazvin irrigation network, in Qazvin province, is located at 150 km-West of Tehran. It is located between 36 20 N, 49 40 E and 36 00 N, 50 35 E. In the present study, a model was developed in the MATLAB programming environment 2017 and linked to the ant colony optimization (ACO) algorithm and the AquaCrop plant growth model. The AquaCrop simulation model is able to assess the crop production under different irrigation water management. Before using the AquaCrop, it is necessary to be calibrated and validated for different plants using appropriate field data at the study’s regional level or regions near to it. For this purpose, the results of the study of Ramezani et al (2017) were used which they used Golkar (1998) for calibration and validation of wheat plant parameters, Farhadi Bansoole (1998) for barley, and Mirlatifi and Sotoodenia (2002) for maize. In addition, the studies of Rahimikhoob et al (2014) were used for forage maize plant and Amiri et al (2017) was used for rapeseed. Extra-heuristic search algorithms, such as the (ACO algorithms, significantly help to solve water resource (Afshar et al., 2015) and management problems (Nguyen et al., 2016). Using ACO can increase the possibility of finding optimal or near-optimal solutions and increase computational efficiency by reducing the amount of search space during the optimization process. Research method was as follow: first, crop, soil and meteorological data from Excel file was exported to the MATLAB code, and the economic productivity for each crop with different irrigation scenarios with connection to Aqua crop model, was calculated and its maximum was determined. Then, by using ACO algorithm and given to water and crop area limitations, optimal crop pattern for all crops under water cultivation of Qazvin irrigation network in three possible water allocation states was obtained so that the network profit would be maximized.
Three different modes of available water volume were considered as A, B and C with supplying 100, 70 and 60% mean of long-term volume of water resources, respectively.. Three different scenarios of depth and irrigation water interval were defined: 1- Depth and similar irrigation interval for autumn and spring crops 2- Depth of different irrigation water and similar interval for autumn and spring crops 3- Depth and different irrigation interval for autumn and spring crops.
The 1, 2 and 3 scenarios were implemented for each of the A, B and C modes. For A, the optimum irrigation schedule was 8-day interval and the irrigation water depths of 25 and 40 mm were determined for autumn and spring crops, respectively. For B, the optimum irrigation schedule was 10-day interval and the irrigation water depths of 24 and 48 mm were determined for autumn and spring crops, respectively. And for C, the optimum irrigation schedule was 8-day interval and the irrigation water depths of 16 and 40 mm were determined for autumn and spring crops. The optimal planting pattern was determined for different scenarios. The maximum area under cultivation in case of 100% water supply was 23029 hectares and by reducing it to 40 and 30% water volume, the area under cultivation will be reduced to 18512 and 17188 hectares, respectively.