A particle swarm optimization (PSO) algorithm is used in this paper for optimal design and operation of irrigation pumping systems. An irrigation pumping systems design and management model is first introduced and subsequently solved with the newly introduced PSO algorithm. The solution of the model is carried out in two steps. In the first step an exhaustive enumeration is carried out to find all feasible sets of pump combinations able to cope with a given demand curve over the required period. The PSO algorithm is then called in to search for optimal operation of each set. Having solved the operation problem of all feasible sets, the total cost of operation and depreciation of initial investment is calculated for all the sets and the optimal set and the corresponding optimal operating policy is determined. The proposed model is applied to the design and operation of a real-world irrigation pumping system and the results are presented and compared with those of a genetic algorithm.

Abstract:Meta-heuristic methods have been widely used for solving complex problems recently. Among these methods, JPSO is regarded as a promising algorithm. However, in order to achieve more robust performance, the probability to solve the graph-based problems is modified by changing the jumping nature of this algorithm and a new algorithm called G-JPSO is presented which is evaluated by solving Fletcher-Powell function and optimal control of pumps in water distribution network problems. In addition to reduction of electricity cost and the problem limitations such as minimum required pressure in each node, minimum and maximum height of tanks, should also be considered. Moreover, another limitation was performed on the objective function which includes the maximum times of turning the pumps on and off. In order to determine the pumps optimal operation, an optimization-simulation model based on the optimization algorithms G-JPSO and JPSO is developed. This proposed model is used for determination of optimal operation program of Van Zyl distribution network. The comparison carried out between the results of our proposed algorithm and those of the similar algorithms including ant colony, genetic and JPSO shows the high ability of the presented algorithm in finding solutions near the optimal solutions with reasonable computation costs.

No abstract