Abstract:

In view of the difficulty in rapidly and effectively detecting the internal short circuitof battery which is the key link in triggering thermal runaway， a short-circuit current estimation method forseries battery pack suitable for different short circuit resistances is proposed in this paper based on the average-difference model. Firstly， the applicability of the average-difference modelfor short-circuit battery and its parameter correction method are analyzed， and the model parameters are identified online by using iterative least squares and Kalman filtering. Then the short-circuit current estimation expression based on model parameters and battery terminal voltage， and the relationship between the instantaneous voltage drop and the short-circuit resistance are derived， and the way of short-circuit current estimation are divided into long-term and short term two methods according to whether there is an obvious instantaneous voltage drop. Finally， the results of verification experiments show that the method proposed can estimate short-circuit current more accurately， and is suitable for short-circuit detection with different resistance values.

Key words: lithium-ion battery, short circuit, fault diagnosis, short circuit current estimation