储能电池外部短路的损伤与失效边界及其预测

The occurrence of external short circuit(ESC) fault of energy storage battery is accompanied by high-rate discharge current, internal heat accumulates rapidly, and battery temperature rises rapidly, which is a typical electro-thermal coupling abuse condition. For ESC of batteries: The characteristics of ESC under different initial conditions were experimentally studied, the time from the start of ESC to voltage and current of battery drop to 0 was defined as ESC failure boundary, and the coupling characteristics of short circuit time and aging are further analyzed. By using electrochemical impedance spectroscopy(EIS), it is determined that the growth of solid electrolyte interphase(SEI) is the main reason for the capacity decline of ESC batteries, which is mainly affected by temperature. Further, taking the time from the start of ESC to the internal temperature of battery reaching the 80 ℃ as damage boundary. To obtain the internal temperature of battery, a lumped parameter thermal model of was constructed, and the internal temperature of battery was estimated based on a PID observer, the estimation error is within 2 ℃. A long-short term memory(LSTM) neural network is constructed to predict the damage and failure boundary of battery. Taking the current, voltage and ambient temperature 1 s before and 2 s after the ESC as input, the results show that the prediction error of damage boundary is within 3.5%, and the prediction error of failure boundary is within 2%. The application of this model can deepen the understanding of battery damage caused by external short circuit, and provide a strong guarantee for battery safety monitoring.