Notebook

Creating a Model¶

In this notebook, we create and solve a single particle model (SPM). This is achieved using a predefined parameter set introduced in Marquis et al. [1]

Before we begin, we need to ensure that we have all the necessary tools. We will install PyBOP and upgrade dependencies.

In [ ]:

%pip install --upgrade pip ipywidgets -q
%pip install pybop -q

import pybop

pybop.PlotlyManager().pio.renderers.default = "notebook_connected"

Note: you may need to restart the kernel to use updated packages.
Note: you may need to restart the kernel to use updated packages.

Creating a Model¶

PyBOP offers the both forward emperical and physics-based forward models. These are provided by PyBaMM, with PyBOP adding wrappers on the underlying classes to reduce complexity and improve stability with parameter inference and design optimisation. Likewise, PyBOP provides a light wrapper on the PyBaMM parameter sets, with user-defined parameters available through the same pybop.ParameterSet class.

Let's construct the parameter set and then the single-particle model (SPM):

In [ ]:

parameter_set = pybop.ParameterSet.pybamm("Marquis2019")
model = pybop.lithium_ion.SPM(parameter_set=parameter_set)

Now that the model is constructed with the Maquis parameter set, we can use the model.predict method as a light wrapper on the PyBaMM.Simulation class. This is the recommended way to generate synthetic data, but not for parameter inference as the performance cost of constructing the Simulation class is high. For parameter inference, model.simulate and model.simulateS1 offer a performant way to solve the forward model with and without sensitivities respectively.

Testing the Model¶

Having constructed the model, we can now have a look at its voltage discharge curve to verify that it is working. The discharge curve is evaluated on the time interval specified by t_eval. model.predict returns the PyBaMM.solution object with all of its functionality. As we are only working with the forward model, PyBaMM plotting methods will be used; however, when performing parameter inference or design optimisation, PyBOP plotting methods are recommended.

In [ ]:

t_eval = [0, 3700]
solution = model.predict([], t_eval)  # No inputs i.e []

# Plot with PyBaMM
solution.plot_voltage_components()

Out[ ]:

(<Figure size 800x400 with 1 Axes>, <Axes: xlabel='Time [h]'>)