BMS records/ processes the data obtained from the battery packs. When these data is compared to the battery model, what's the maximum acceptable accuracy/RMSE in terms of mV or % ? Additional literature or reference might help me
Practically, the best commercially-available ICs for this application have a typical voltage accuracy values quoted as +/- 0.2mV , with worst-case accuracy of 1 to 2mV (e.g. Linear LTC6811 which uses a 16bit ADC). Low-cost consumer electronic applications would typically use devices with poorer accuracy than this, though.
From a requirements point of view, cell voltage measurement accuracy has a strong bearing on:
- SOC estimation accuracy- particularly for Li FePO4 cells with a flat discharge curve
- cell balancing accuracy
- it's also used for monitoring cell safety- here it's also the speed of detecting an over or under voltage that is important, as well as the accuracy.
Assuming you are asking with reference to LFP cells and with reference to the figure attached, If the OCV-SoC curve measured at 1 h after discharge is used as the reference to estimate the SoC from a voltage measurement, the voltage evolution during relaxation between 1 h and 24 h at 45% SoC (from Point B to Point C) will introduce an error of about 4% into the SoC estimation [1].
Therefore with regards to the flat discharge curve of LFP cells, I’d propose a measurement accuracy of at least 1 mV. Even 2 mV is, according to me, not a good enough option.
Regards,
Vipul Dhingra
[1] Li A, Pelissier S, Venet P, Gyan P. Fast Characterization Method for Modeling Battery Relaxation Voltage. Batteries. 2016; 2(2):7.