This is my first post. In electrochemistry, complex impedance data is modelled using equivalent circuit models. This is based on non-linear least squares fitting of data comprising of real and imaginary complex function. Typical plot recorded at different conditions is attached. I wish to know if TensorFlow could be used to model such experimental data keeping in mind the actual electrochemical processes occurring in the system, which is a pre-condition for selecting an equivalent circuit model. Input is real impedance and complex impedance and the output should be the fitted data. In general, I wish to know if multiple sets of data (as shown in the image) can be simultaneously fitted using a single ANN model as ir pertains to a single system but under different conditions. The plot shown is at a particular temperature (773 K). I can add data sets recorded at other temperatures as well so that we can have a generalized ANN model for a specific system. Is it possible?
Sorry for the delayed response. For your use case, the comple impedance can be modeled using ANN. The model can be fit with multiple sets of data(real Impedance, img impedance and temperature). You can model real impedance, img impedance and temperature as the input and complex impedance as a target variable using either numpy or pandas libraries and then can feed to your model. CNN models also suggestible for your use case. Here is the sample guide for further assistance.
Thank You