Identifying Low-Order Frequency-Dependent Transmission Line Model Parameters

This paper describes the modeling and parameter identification of a frequency dependent transmission line model from time-domain data. To achieve this, a single-phase transmission line model was implemented in OpenModelica where the frequency dependent behavior of the line was realized by a series of rational functions using the Modelica language. Next, the developed line model was exported as a Functional Mock-up Unit (FMU). The RaPId toolbox was then used for automated parameter optimization of the model within the FMU that was interfaced to RaPId via the FMI Toolbox for MATLAB. Given a reasonable starting guess of the set of parameters, the toolbox improved the model's response significantly, resulting in a good approximation even though low-order representations were used for the identification process. It was found that even though the process was straightforward, it can be enhanced by exploiting the physical/numerical properties of this specific problem.