Diagnostic method for PEM fuel cell states using probability Distribution-Based loss component analysis for voltage loss decomposition.

• This study proposes a novel method referred to as the loss component analysis (LCA) to represent the current state of fuel cells. • LCA is a diagnostic method derived from independent component analysis. • The proposed fuel cell diagnosis method analyzes the fuel cell state using probability density functions. • The method can diagnose a fuel cell state by obtaining weight parameters assigned to each loss component. • Consistent diagnostic results and interpretations were possible for simulation and all the experimental data. This study proposed a novel method referred to as the loss component analysis (LCA) to represent the current state of fuel cells. The LCA method was derived from an independent component analysis (ICA) and used probability density functions of activation, ohmic, and concentration losses. This method determined three weights related to each loss component reflecting the fuel cell states, and the fuel cell conditions were diagnosed using deviations in weight from the reference weight at the normal state. The maximum increase in weight allocated to each loss component was found to have the most significant impact on changes in the state of the fuel cell from its normal state. Moreover, LCA was applied to both the data obtained from empirical models and the data acquired through experiments that mimic the three faults that could occur during fuel cell operation. The results were compared to demonstrate the validity of the proposed method. [ABSTRACT FROM AUTHOR]