Unit Sizing of a Stand-Alone Hybrid Power System Using Model-Free Optimization

Abstract

In this paper a model-free optimization method is applied to the problem of unit sizing in a hybrid power system such that demand of residential area is met. Optimal sizing of two systems is considered. In the system No.l, the produced power is delivered to the load and the hydrogen produced by the reformer is stored in the tank. If the power produced by the wind turbine is more than the demand, the remainder of wind turbine's power is delivered to the electrolyzer to produce hydrogen, such that when the wind power cannot meet the demand, the fuel cell is fed by the stored hydrogen and produces enough power, together with the wind turbine's power. In the system No.2, the hydrogen produced by the reformer is delivered to the fuel cell directly. When the power produced by the wind turbine plus power produced by the fuel cell (fed by the reformer) is more than the demand, the remainder is delivered to the electrolyzer. In contrast, when the power produced by the wind turbine plus that produced by the fuel cell (fed by the reformer) is less than the demand, some more fuel cells are employed and they are fed by the stored hydrogen. Our aim is to minimize the costs of the system such that the demand is met. PSO algorithm is used for optimal sizing of system's components.