| 초록 |
It has been well known that the performance and durability of a PEM (polymer electrolyte membrane) fuel cell system is strongly affected by the relative humidity of gaseous streams flowing inside the system. Thus, the accurate prediction of the relative humidity is essential for monitoring and controlling the performance and durability of a fuel cell system. In this study, three different types of models were developed for predicting the relative humidity: 1) an empirical model based on artificial neural network, 2) a mechanistic model, and 3) a hybrid model combining the empirical and mechanistic models. The three different modes were applied to a 30 kW-class PEM fuel cell system, and the predictive performances of them were compared with each other. Among the models, the empirical model showed the best predictive performance. The developed models can be used as soft sensors replacing the actual humidity sensors in PEM fuel cell systems. |
