Optimal sizing of a hybrid power system considering wind power uncertainty using PSO-embedded stochastic simulation

Anticipated high penetration of stochastic energy flows throughout the stand-alone micro grids should be optimized by using hybrid stochastic-heuristic simulation methods. It is well treated, in this way, both the uncertainty caused by the renewable power production and the non-linearity of the objective function. In this paper, a hybrid simulation procedure is employed to the problem of sizing in a hybrid power system considering wind power production uncertainty. The developed algorithm consists of a particle swarm optimization (PSO) subroutine embedded in a multivariate Monte Carlo simulation. This study is performed for Kahnouj area in south-east Iran. The system consists of fuel cells, wind turbines, some electrolyzers, a reformer, an anaerobic reactor and some hydrogen tanks. The system is assumed to be stand-alone and uses the biomass as an available subsidiary energy resource. The main objective is to minimize the total costs of the system in view of wind power uncertainty to secure the demand. PSO algorithm is used for optimal sizing of system´s components for each simulation run used by Monte Carlo method. Besides, several statistical modeling and analyses are performed prior to the simulation and later on to properly interpret the results.