A novel method for equivalent circuit fitting in python using differential evolution

Electrochemical impedance spectroscopy (EIS) is an important tool for analyzing membrane resistance in fuel cells, as well as separating different processes according to their time domain. Measurement results can be described using parameters through the modelling of equivalent circuits, which can take hours to days, depending on the settings used. An equivalent circuit takes the complex processes happening inside a fuel cell and tries to find and electrical circuit, that represents the electrical characteristics of a fuel cell. The objective of this work is to simplify and shorten the equivalent circuit fitting using a differential evolution algorithm with the premise of obtaining a good match between data and model.
Differential evolution is used to find the global minimum of multivariate functions like the ones used to describe equivalent circuits. Using differential evolution, we were able to avoid premature termination of the optimization due to a local minimum in the multivariate function. After optimizing the method with more than a billion combined evolutions, the required computational time was reduced significantly, to under 10 minutes, while managing to retain agreement between the resulting model and measured data. The optimization showed a significant decrease in the computational time required compared to the default, with only a slight change in the result.
These results will allow us to implement the equivalent circuit fitting in a data analysis routine to compare equivalent circuit parameters over the course of measurements. The saved computing time can also be used to go beyond the representation with electrical components and introduce advanced circuit models such as dynamic large scale equivalent circuits (dLSEC) into the data analysis process.