Modelling and analysis of an original direct hybridization of fuel cells and ultracapacitors

详细信息    查看全文

• 作者：C. Turpin^a ; ^b ; ^c ; ^{christophe.turpin@laplace.univ-tlse.fr" class="auth_mail" title="E-mail the corresponding author} ; D. Van Laethem^a ; ^b ; ^c ; B. Morin^a ; ^b ; ^c ; O. Ralliè ; res^a ; ^b ; ^c ; X. Roboam^a ; ^b ; ^c ; O. Verdu^d ; V. Chaudron^d
• 关键词：Fuel cell ; Ultracapacitor ; Direct hybridization ; Large signal modelling ; Parameterization
• 刊名：Mathematics and Computers in Simulation
• 出版年：2017
• 出版时间：January 2017
• 年：2017
• 卷：131
• 期：Complete
• 页码：76-87
• 全文大小：2478 K

文摘

The feasibility of fuel cell applications has been demonstrated throughout the world illustrating all the potentialities of this technology. Research efforts are currently focused on improving life time and reducing costs. Power variations imply fluidic variations for a fuel cell system. Because of a non-infinite dynamic of the gas supplies, bad local conditions (low gas concentrations, pressure stresses, water accumulation, etc.) can occur within the electrodes degrading more or less their lifetime. This phenomenon is increased in the case of a Hsource">₂${}_2$/air fuel cell because of the relatively slow response time of the air compressor. These conditions are mostly created by the application of severe and frequent load peaks or by the interactions with power converters’ current harmonics. To reduce this effect, hybridization with an electrochemical storage component (typically ultracapacitors) is generally suggested via one or two power converters, requiring the implementation of an energy management. Moreover, in most of these architectures, the fuel cell is not protected against the current harmonics generated by its own power converter. Here is proposed a structure called direct hybridization where fuel cells and ultracapacitors are directly associated at the elementary scale, permitting the double protection sought, limiting the interactions with the power electronics and providing a natural energy management (no external control required). Firstly, the authors will establish a large signal model for this original system. Secondly, they will analyse certain properties of this integrated component in terms of dynamic capabilities under current steps through comparisons between simulations with experiments.