Simple Model and Experimental Identification of a Fuel-Cell-Based Power Supply Oriented to System-Level Analysis

Power electronics systems are becoming larger and more complex, involving a wide variety of loads, power sources, and power conditioning equipment. Simple behavioral models are required to perform a system-level analysis. These models can be parameterized from simple measurements or data supplied by the manufacturer. One of the emerging power sources are the fuel cells, which are being used in different power levels and applications. Some commercial products include not only the fuel cell stack, but also a dc-dc converter to provide a regulated output voltage despite the operation conditions of the fuel cell. In this paper, a behavioral model of a commercial fuel cell with a built-in front-end dc-dc converter is developed, based on load step tests. Model identification is carried out from experimental measurements using well-known identification algorithms. A validation of the obtained model has been performed by means of applying the same current patterns to the model and the actual fuel cell: steps; pulses; and sinusoids. Moreover, a simple power system has been built in order to validate the proposed model. A good trade-off between simplicity and accuracy is achieved.