RC circuit simulation data and look up table mismatch. Help!

Dear all,

I have an issue with a mismatch in simulation data. I generated simulattion using below RC circuits, but the generated data does not match Look up table!

openExample('simscape/LithiumBatteryCellOneRCBranchEquivalentCircuitExample')

I am working on this RC circuit simulation for HPPC test to extract parameters and I have 2 questions regarding on this model. I am new to this field so much appreciate for your help!

1. I used the default setup of this model, without changing anything. In this model, LUT tables already have data. For example, at T = 20 + 277.15, SOC =1, R0 = 0.0085. I run the model with step size 1, 100 and fix step size 1. At the first discharge pulse, I used Voltage before and After the ohmic drop to calculation internal resistance R0. The result is R0= 0.00861 for all 3 cases. This results seem close ( 0.00861 vs 0.0085) but I really need an accurate results because R0 at different SOC levels are very similar. How can I adjust the model to get the same number as LUT?

This is Ro with variable step size 1

This is R0 with fix step size 1

2. The results is close for R0, but for R1. It was totally different. Correctly me if I was wrong because I am new to this. There are different sources shows how to calculate R1. From my understand, I can get R1 from the voltage different in Polarization process or Voltage relaxation process, then use curve fitting to find constant time tau, then C1 = tau/R1. However, in both cases, I got bad results that not even close to R1_LUT. How can I can make it right?

And again, I used the the default model above, without changing any thing, I only used variable step with max step size = 1, 100 anf fix step size 1 in model settings solver

Thank you so much in advance!

Answer 1

[Thanh]

Any know? Please help!

Answer 2

[Sameer]

Hi @Thanh

The small difference in R0 (0.0085 Ω in LUT vs. ~0.00861 Ω measured) is normal. It’s caused by solver step timing, numerical tolerances, and possible interpolation in the SOC/temperature lookup. The battery block updates R0 each step based on current SOC and T, so even tiny changes will shift it slightly.

To match LUT more closely:

• Use a smaller fixed step or tighter solver tolerances (RelTol ≤ 1e-5).
• Keep temperature constant (disable thermal dynamics).
• Capture voltage exactly at the current step change (use hit-crossing or fine time step).

For R1 extraction:

• After the pulse, fit the voltage relaxation to an exponential:
• Compute , then .
• Use a long enough rest period and smooth data before fitting.

Even with best practices, expect ≤0.1 % deviation from LUT due to solver integration and interpolation that’s within the model’s numerical accuracy.

Hope this helps!