How to simulate multiple drive cycles? (using the modulo function)

[anandmy]

I am trying to simulate a drive cycle for a long time duration (at least 1 year). Currently I have a drive cycle representing a single day's current profile and want to replicate it for multiple days (years hopefully). While this is easier to imagine when using the Experiment class as it allows using lists, is there a way of doing the same for the Current Function [A] parameter or using a different approach?

Tino suggested using the modulo function (t % period) but I could not figure out how to use that. Any help would be appreciated. Thank you.

[brosaplanella]

You first need to define the interpolant from your dataset. You can find an example of how to do so in the "Loading current data section" of the notebook about how to change the input current.

Note that when you create the interpolant

current_interpolant = pybamm.Interpolant(drive_cycle, timescale * pybamm.t)

you pass as the variable the dimensional time timescale * pybamm.t. In your case, you want to take the dimensional time modulo the period. In your case, the period is the duration of the drive cycle, so you can probably read that from the data as well. Assuming that you have already defined the period you can then write

current_interpolant = pybamm.Interpolant(drive_cycle, (timescale * pybamm.t) % period)

Note that you can call the modulo operator both by using % and pybamm.Modulo(left, right). See the documentation for more information.

[valentinsulzer]

The following example repeats the US06 drive cycle 3 times (for 1800s instead of 600s):

import pybamm
import pandas as pd
import os

os.chdir(pybamm.__path__[0] + "/..")

pybamm.set_logging_level("INFO")

# load model and update parameters so the input current is the US06 drive cycle
model = pybamm.lithium_ion.SPMe({"thermal": "lumped"})
param = model.default_parameter_values


# import drive cycle from file
drive_cycle = pd.read_csv(
    "pybamm/input/drive_cycles/US06.csv", comment="#", header=None
).to_numpy()

# create interpolant
timescale = param.evaluate(model.timescale)
# in the following line we use % 600 to repeat the drive cycle every 600 seconds
current_interpolant = pybamm.Interpolant(drive_cycle, (timescale * pybamm.t) % 600)

# set drive cycle
param["Current function [A]"] = current_interpolant


# create and run simulation using the CasadiSolver in "fast" mode, remembering to
# pass in the updated parameters
sim = pybamm.Simulation(
    model, parameter_values=param, solver=pybamm.CasadiSolver(mode="fast")
)
# we need to specify the solve time to be different from the default [0, 600]
# this will raise a warning but this can be ignored
sim.solve([0, 1800])
# use this instead to get rid of the warning
# t_drive = drive_cycle[:, 0]
# step_size = np.min(np.diff(t_drive))
# t_eval = np.arange(0, 1800, step_size)
# sim.solve(t_eval)
# you can play with step_size to change the level of resolution of the drive cycle

sim.plot(
    [
        "Negative particle surface concentration [mol.m-3]",
        "Electrolyte concentration [mol.m-3]",
        "Positive particle surface concentration [mol.m-3]",
        "Current [A]",
        "Negative electrode potential [V]",
        "Electrolyte potential [V]",
        "Positive electrode potential [V]",
        "Terminal voltage [V]",
        "X-averaged cell temperature",
    ]
)

[ssingh-ipa]

sim.solve(starting_solution=sol)
https://github.com/pybamm-team/PyBaMM/blob/develop/examples/notebooks/simulating-long-experiments.ipynb

Looks like that starting_solution is only valid for experimental inputs and not for driving cycles.

Also, executing https://github.com/pybamm-team/PyBaMM/blob/develop/examples/notebooks/simulating-long-experiments.ipynb currently gives me the following error -
"esoh_model = pybamm.lithium_ion.ElectrodeSOH()
AttributeError: module 'pybamm.models.full_battery_models.lithium_ion' has no attribute 'ElectrodeSOH'"

[valentinsulzer]

Also, executing https://github.com/pybamm-team/PyBaMM/blob/develop/examples/notebooks/simulating-long-experiments.ipynb currently gives me the following error -
"esoh_model = pybamm.lithium_ion.ElectrodeSOH()
AttributeError: module 'pybamm.models.full_battery_models.lithium_ion' has no attribute 'ElectrodeSOH'"

make sure you are using the develop branch of pybamm if testing examples on the develop branch. If you installed with pip look for examples on the main branch

[asifsultan-THD]

I want to simulate it for drive cycle but in the above lines I am facing issue

timescale = param.evaluate(model.timescale)
current_interpolant = pybamm.Interpolant(drive_cycle, timescale * pybamm.t)

---

TypeError Traceback (most recent call last)
C:\Users\AAA27~1.SUL\AppData\Local\Temp/ipykernel_2976/286842135.py in
2 timescale = param.evaluate(model.timescale)
3 in the following line we use % 600 to repeat the drive cycle every 600 seconds
----> 4 current_interpolant = pybamm.Interpolant(drive_cycle, timescale * pybamm.t) % 600)

TypeError: init() missing 1 required positional argument: 'children'

[valentinsulzer]

In the latest versions, you have to do Interpolant(data[0], data[1], timescale * pybamm.t)

[asifsultan-THD]

Thanks for the reply. Actually I am stuck in one of your more line. Its getting simulated. But its not getting solved and giving runtime error. I am talking about this line:------ sim.solve([0, 1800])
Using the code below this line also giving the same error

And the error is :- RuntimeError: .../casadi/core/function_internal.cpp:145: Error calling BSplineInterpolant::init for 'LUT':
.../casadi/solvers/bspline_interpolant.cpp:91: Assertion "x.size()>=2*m+2" failed: Need more data points

Please correct it just been fews days on Pybamm
Thanks in Advance

[Gokarna-Bahadur-Thapa]

Help me to correct this code

create interpolant

timescale = param.evaluate(model.timescale)

in the following line we use % 600 to repeat the drive cycle every 600 seconds

current_interpolant = pybamm.Interpolant(drive_cycle, (timescale * pybamm.t) % 600)

NotImplementedError Traceback (most recent call last)
Cell In[10], line 2
1 # create interpolant
----> 2 timescale = param.evaluate(model.timescale)
3 # in the following line we use % 600 to repeat the drive cycle every 600 seconds
4 current_interpolant = pybamm.Interpolant(drive_cycle, (timescale * pybamm.t) % 600)

File ~\anaconda3\Lib\site-packages\pybamm\models\base_model.py:294, in BaseModel.timescale(self)
292 @Property
293 def timescale(self):
--> 294 raise NotImplementedError(
295 "timescale has been removed since models are now dimensional"
296 )

NotImplementedError: timescale has been removed since models are now dimensional

[agriyakhetarpal]

Here is a related discussion about how to adapt your code around the removal of timescale due to dimensional models: #3712