How to fit a model to experimental data

[manishkothakonda]

I am attaching a screenshot of the data I am working on here, which is constant discharge data of a 10Ah battery from 100% SOC to 0% SOC for a 0.1C discharge rate for a temperature of 40°C

I have other data in the spreadsheet related to 10Ah battery from 100% SOC to 0% SOC for a 0.2C, and 1C discharge rate for a temperature of 40°C, 25°C, 5°C, and -25°C,

I would like some assistance on what path could I take to fit this data into a model, and the parameters to optimize.
I would like to evaluate the fitted battery model to 1C constant current discharge from 100% SOC at an ambient temperature of -25°C for the following time periods in seconds: t = 0, 20, 40, 60, 80, 100, …, 3000, 3020, 3040, 3060, 3080, 3100

This is my first time using PyBaMM and I am an adept user of Python.

[brosaplanella]

You might want to try pybamm-param, which is a package for fitting PyBaMM models to data. Note that the package is under development so is neither stable nor well-tested at the moment.

We can not advise on which parameters to fit, as this is a research question.

[manishkothakonda]

Thanks, @brosaplanella, I appreciate you pointing out the package. However, I am unable to install the pybamm-param package. I tried installing in mac os following the exact commands. The error I see is when I do "pip install -e .[dev]", which replies back as "zsh: no matches found: .[dev]"

Are there specific guidelines that you could share to help me install this python package?

Thank you very much

[brosaplanella]

Try installing directly from Github: pip install git+https://github.com/paramm-team/pybamm-param

[manishkothakonda]

Thanks, This works!

[agriyakhetarpal]

Hi, the pattern matching for zsh-based shells is different from that of bash – you should use pip install -e .'[dev]' instead (enclosing it with single quotes) and it should work :)