Optimal delivery and distribution of water in the BP14 canal of Foumanat irrigation network using optimization algorithms PSO and ICA

Canal systems of distributing water have different design capacities, command areas and the lengths requiring different duration of operation. Irrigation scheduling becomes a complex process under these conditions, especially for rotational water distribution. Planning for optimal operation in irrigation network is an efficient solution in water saving, increasing in cultivation and improving agriculture performance. Delivery and distribution in irrigation network are multi-objective and have multi variation optimization and multi limitation problems. By the optimal delivery schedule, minimum canal capacity and minimum delivery time is derived. Minimization of canal capacity reduces the construction costs and minimization of delivery time, limits water losses due to seepage, evaporation and water stealing. Solving such problem needs powerful optimization methods. Evolutionary algorithms are the major group of search algorithms for finding a suitable answer, using current evolutionary rules in nature. In this research ICA and PSO algorithms are applied. To achieve the research goals, the considered algorithms are applied in MATLAB 2014 software. The results of this research are also compared with GA method in study of Monem et al, (2007) and Ants Colony System Algorithm in study of Emadi and Kakouie (2012).
Imperialism is the policy of extending the power and rule of a government beyond its own boundaries. ICA is a novel global search strategy that uses imperialism and imperialistic competition process as a source of inspiration. This algorithm is based on that in real world countries try to extend their power over other countries to use their resources and bolster their own government. In fact, imperialist countries attempt to dominate other countries and turning them to their colonies. Also, imperialists compete strongly with each other for taking possession of other countries. During this competition stronger empires get more power and the weakest one ultimately collapses. This general policy of imperialist competition is used as the basis of the ICA. Particle swarm optimization (PSO) is a population-based stochastic approach for solving continuous and discrete optimization problems. In this research, the delivery and distribution program in distribution channel branches with ICA and PSO algorithm were provided so that the various objects(such as decreasing in distributor channel capacity and decreasing in time) needed for complete the irrigation program optimize, as a single and two objectives. In this program, defined inputs to the model were: first branch numbers, the upper and lower limit of delivery discharge to each branch and the branch coverage, gross irrigation requirement, irrigation interval and block numbers. By running the model, the best intermittent of branches in per block, minimum of distributer channel capacity and minimum irrigation duration in the optimum conditions define to the model, as outputs. In order to comparison ICA and PSO result with ACS and GA, the model is applied on BP14 canal in Fomanat Irrigation Network in west Gilan Province, Iran. This canal has a length of 6852 meters with a trapezoidal cross section and capacity of 2.5 m3/s and land cover area of 2100 ha. Irrigation water requirement is 50 mm.
The optimal delivery schedule canal for single objective (canal capacity minimization) is derived as Follow: canal capacity: 595 lit/s, time delivery: 343hr and the number of upstream gate regulations: 12 in PSO algorithm; and canal capacity: 546 lit/s, time delivery: 357hr and the number of upstream gate regulations: 11 in ICA. The optimal delivery schedule canal for two objective including minimization canal capacity and delivery time is derived as follow: canal capacity: 645lit/s, time delivery: 300hr and the number of upstream gate regulations: 12 in PSO algorithm; and canal capacity: 630lit/s, time delivery: 292hr and the number of upstream gate regulations: 12 in ICA.
ICA had the better Optimal Water Scheduling comparing with GA, ACS and PSO. Comparison of results obtained from ICA with other method's result showed that ICA in both single and two objective state determined less canal capacity. Also, in ICA the number of upstream gate regulations was one less than other approaches. It was found that PSO and ACS had the best performance in the next level. So, ACS determined less canal capacity and PSO determined more irrigation schedule complement duration. Therefore, ICA algorithm is a powerful method that able to escape from the local peaks and reach to global optima. ICA algorithms more likely reach Global optimum positions than PSO and GA. This algorithm, due to wide space search, could create a better collocation for branches of canal in irrigation network which leads to the better optimization.