How to correctly estimate simulation time and error in a battery cell simulation

I am modeling a battery cell using an equivalent circuit model. I have 3 different models for 1RC, 2RC and 3RC. However when i measure the simulation time using the profiler i get similar times with all three circuits. Morover the error associated with the simulation also stays roughly the same. According to the bibliography the more RC branches the circuit has, the longer it takes to simulate but the less error it has.

Anyone has any idea where this problem might be comming from?

The circuit uses LUTs for the circuit parameters and a SoC estimation block(coloumb counting)

Thanks.

Answer 1

[Javier Gazzarri]

Hello Jorge,

Your statement about error and simulation time is correct. However, if the model is simple (e.g. it only has a single cell) you will not notice a significant performance difference because MATLAB will be able to run it very fast in both cases.

The reason why you are not seeing a difference in accuracy may be that your simulation is mainly exciting the first order in the ECM dynamics, so the other two RC pairs are not contributing significantly. This is the case, for example, when the second and third time constants are large compared to the dynamic behavior of your input. For example, if the current is changing on a second-time-frame, time constants of the order of 100 or 1000 seconds will not play a significant role in the simulation.

Please let me know if this helps.

Javier