Identifiability of Generalized Randles Circuit Models

The Randles circuit (including a parallel resistor and capacitor in series with another resistor) and its generalized topology have widely been employed in electrochemical energy storage systems, such as batteries, fuel cells, and supercapacitors, also in biomedical engineering, for example, to model the electrode-tissue interface in electroencephalography and baroreceptor dynamics. This paper studies identifiability of generalized Randles circuit models, that is, whether the model parameters can be estimated uniquely from the input-output data. It is shown that generalized Randles circuit models are structurally locally identifiable. The condition that makes the model structure globally identifiable is then discussed. Finally, the estimation accuracy with respect to noise-free, noisy, zero-mean, and nonzero-mean data is evaluated through extensive simulations. The existing tradeoff between the estimation of Warburg term and other parameters by using zero- and nonzero-mean data is fully discussed.