A Comparative Analysis Between Genetic Algorithms and Complex Nonlinear Least Squares on Electrical Impedance Characterization

This paper compares the performance of the complex nonlinear least squares algorithm implemented in the LEVM/LEVMW software with the performance of a genetic algorithm in the characterization of an electrical impedance of known topology. The effect of the number of measured frequency points and of measurement uncertainty on the estimation of circuit parameters is presented. The analysis is performed on the equivalent circuit impedance of a humidity sensor.

Vector Fitting based Automatic Circuit Identification

This paper describes a method to automatically obtain, from a set of impedance measurements at different frequencies, an equivalent circuit composed of lumped elements based on the vector fitting algorithm. The method starts from the impedance measurement of the circuit and then, through the recursive use of vector fitting, identifies the circuit topology and the component values of lumped elements. The method can be expanded to include other components usually used in impedance spectroscopy. The method is firstly described and then two examples highlight the robustness of the method and showcase its applicability.