Why a battery simulation is needed?

Model definition -> Environment and experiment definition -> Simulation -> Insight -> Well prepared Battery Design

Producing a battery will cost some money. a bad design or unfit design of battery may ended up in losing money. Hence, it is really important to perform simulation before actual design to prevent unfit battery. Each battery will be designed based on specific purpose and condiition. Simulation will help to simualte the battery behavior during specific environment/test and researcher can use the simulation result to prepare the actual design design.

Pybamm as Battery Simulation Tools

Currently there is one growing open-source python package that specifically function to simulate a battery. The package name is pybamm. Its a tools to simulate battery using mathematical approach.

Pybamm also has built in integrated data science package like sciPy, numPy, etc.

Based on their testimonies, pybamm is used by battery researcher for educational or industrial purposes. Since the code is open source (available to the public), all researcher over the world can help to enhance the package.

Installing PyBamm.

Pybamm is python package, its advised to use specific python version : 3.12.x. Newer version of python might not be able to run pybam as its require lower version of scipy that can only run under python 3.11.x or 3.12.x

Creating python package for PyBam using MiniConda

create a new project using python 3.12 or 3.11 with MiniConda/AnaConda. used MiniConda so all the installed package can be fully defined from scatch. Anaconda has prebuilt data science packge (tensorflow, numpy, etc). MiniConda is more streamlined.

conda create --name sdbms-pybamm python=3.11

if project creation is success, it will produce following folder skeleton.

postimage100Miniconda Project Skeleton

active the miniconda virtual environment

    conda activate <project_name>

Install pre-requisite package

Apart from pybamm itself, other python library will be used

  • Pybamm
  • Numpy
  • Jupyterlab
  • SciPy
  • Pandas

Install jupyterlab

its better to use jupyterlab so that notebook can be used (.ipynb) executing data science project. Every procedure can be documented, listed and arranged properly. It gave a flexible working environment compared to normal python script.

conda install -c conda-forge jupyterlab

ensure the jupyterlab installed at the conda environment.

conda list --name <project_name> | grep jupyter

Install Pybamm, Numpy and Skipy

below command is used to install Numpy, SciPy and pandas

conda install -c conda-forge scipy numpy pandas

ensore those package installed

conda list --name <project_name> | grep scipy
conda list --name <project_name> | grep numpy
conda list --name <project_name> | grep pandas

Install ipywidget for the plot

execute following command

conda install -c conda-forge ipywidgets

validate

conda list --name <project_name> | grep ipywidgets

Install plotly to enable the plot slider

execute following command

conda install -c conda-forge plotly

validate

conda list --name <project_name> | grep ipywidgets

add to the cell

!jupyter nbextension enable --py widgetsnbextension

execute via terminal

jupyter labextension install @jupyter-widgets/jupyterlab-manager

Generate environment.py

execute below script to generate environment details (for future installation by different people)

conda env export --from-history > environment.yml

Full Code of Experiment

Code is written using ipynb, all instruction, available model, input model param, model output are written on the code.

Github Repo

repository for those experinment can be found here:

https://github.com/deganandapriyambada/battery-simulation-pybamm/blob/master/notebook/battery.simulation.ipynb

Steps to replicate

Installation Method

conda create --name <project_name> python=3.11
conda env create -f environment.yml
conda activate <project_name>