Primary works about optimization have investigated selection of a suitable model. Broka and Ekdunge were among the first researchers to model the catalyst layer (Broka and Ekdunge 1997). Their results showed that the agglomerate type model had the best agreement with empirical results compared to the pseudo homogenous model, especially at high current density. Also Sui et al. modeled and optimized the catalyst layer and confirmed the above-mentioned results (Sui et al. 1999). Following surveys have evaluated parameters of the catalyst layer. For instant, Wang et al. studied the effect of structural parameters in the catalyst layer and determined oxygen diffusion coefficient, proton conductivity, and agglomerate size as key parameters (Wang et al. 2004). Furthermore, Sun et al. accounted for catalyst layer thickness, catalyst loading, and ionomer content as the most important parameters (Sun, Peppley and Karan 2005). Following that, other studies have focused on optimization of objective functions. For example Srinivasarao et al. researched optimization of the catalyst layer in terms of current density and cost of the catalyst layer separately as two objective functions. They reduced the cost power ratio by about 40% and improved performance by about 10% (Srinivasarao 2011). Also Kulikovsky et al. presented an optimized model for the catalyst layer and showed that optimal catalyst distribution had a more important effect than optimal ionomer distribution at high current density (Kulikovsky 2012). Recent studies have been directed toward multi objective optimization of fuel cell systems. For example Mert, Feali and Ang et al. conducted multi objective optimization for direct methanol fuel cell (Mert and Özçelik 2013), microfluid fuel cell (Feali and M. Fathipour 2014), and PEM fuel cell systems (Ang, Brett and Fraga 2010), respectively. Hereafter, studies about multi
Primary works about optimization have investigated selection of a suitable model. Broka and Ekdunge were among the first researchers to model the catalyst layer (Broka and Ekdunge 1997). Their results showed that the agglomerate type model had the best agreement with empirical results compared to the pseudo homogenous model, especially at high current density. Also Sui et al. modeled and optimized the catalyst layer and confirmed the above-mentioned results (Sui et al. 1999). Following surveys have evaluated parameters of the catalyst layer. For instant, Wang et al. studied the effect of structural parameters in the catalyst layer and determined oxygen diffusion coefficient, proton conductivity, and agglomerate size as key parameters (Wang et al. 2004). Furthermore, Sun et al. accounted for catalyst layer thickness, catalyst loading, and ionomer content as the most important parameters (Sun, Peppley and Karan 2005). Following that, other studies have focused on optimization of objective functions. For example Srinivasarao et al. researched optimization of the catalyst layer in terms of current density and cost of the catalyst layer separately as two objective functions. They reduced the cost power ratio by about 40% and improved performance by about 10% (Srinivasarao 2011). Also Kulikovsky et al. presented an optimized model for the catalyst layer and showed that optimal catalyst distribution had a more important effect than optimal ionomer distribution at high current density (Kulikovsky 2012). Recent studies have been directed toward multi objective optimization of fuel cell systems. For example Mert, Feali and Ang et al. conducted multi objective optimization for direct methanol fuel cell (Mert and Özçelik 2013), microfluid fuel cell (Feali and M. Fathipour 2014), and PEM fuel cell systems (Ang, Brett and Fraga 2010), respectively. Hereafter, studies about multi