Multi-impedance Distribution of Relaxation Times Applied to Predicting Fuel Cell Stack Operating State: A Theoretical and Experimental Study

The distribution of relaxation times (DRT) converts the impedance data into a simple peak signal, which can be used to quickly identify the operating state of the fuel cell. However, the DRT peak obtained from a single impedance value cannot accurately define the internal reaction process in the fuel cell. In this paper, the variation pattern of the operating state of the fuel cell stack is determined by analysing the peaks of each single cell. The results show that, for DRT, a regularization parameter (lambda) of 10-5 can produce the optimal peak, whereas the change in the shape of its peak can be better used to analyse the operating state of the stack. Moreover, the operating state of the stack can be adjudged and predicted by analysing the changes of P1, P2 and P3 peaks of DRT. Additionally, the reasons for lower performance were also analysed.