Towards in-situ detection of nascent short circuits and accurate estimation of state of short in Lithium-Ion Batteries

Isolation of relaxation times under open-circuit conditions: Toward prognosis of nascent short circuits in Li-ion batteries

“…5 , 6 , 7 While many SC detection protocols have been proposed in the literature, due to many of their handicaps like lack of early detection, dependence on huge, specific datasets, inaccuracy and difficulty in implementation, etc., they fail to meet the requirements for BMS deployment. 8 , 9 In this work, we attempt to solve these problems. At the core of the proposed approach lies the understanding that the LIB experiences a voltage recovery after charge/discharge current is interrupted, due to relaxation of concentration profiles developed inside solid electrode particles.…”

“…According to one industry standard (GB/T 31484-2015), the maximum leakage current allowed in a battery system to be classified as a soft short is C/3.7 (where C is the 1C current). 9 , 15 , 16 This translates to a short resistance, R sh , of approximately 3Ω–6Ω for typical commercial mobile phone batteries (example: for a 4Ah battery with nominal voltage ∼3.85V, R sh = [3.85V/(4A/3.7)] ∼ 3Ω), with this number varying, depending on the battery capacity/form factor. An easier way to differentiate the soft-hard transition is the short resistance below which the signatures become noticeable.…”

“…However, as the short severity is low (high R sh ), its influence on the battery states is very subtle to be noticed, making reliable detection hard. 9 , 15 Not only do proposed methods in the literature not detect soft short but they also pose a few other concerns.…”

“…Recently, a pulse current-based technique was shown to work excellently in detecting soft short, as soft as 200Ω. 9 However, this method requires high-amplitude currents that may not be allowed by the power management integrated circuit (PMIC) onboard different battery applications, apart from the fact that an onboard ECT model was also necessary to monitor the states and accordingly calculate the pulse current required. However, apart from the practicality aspect of these algorithms, one of the major handicaps is their insensitivity toward detecting early-stage SCs, which is ideally when they should be detected.…”

Fault Diagnosis Method for Lithium-Ion Battery Packs in Real-World Electric Vehicles Based on K-Means and the Fréchet Algorithm

“…The current research on battery for electric vehicles has been mentioned in many types of literature, such as battery fault diagnosis, estimation of remaining useful life for batteries, state of the health estimation, etc. − And the research approaches in the literature about fault diagnosis can be broadly classified into three categories: knowledge-based, model-based, and data-driven fault diagnosis approaches. − Among them, the knowledge-based fault diagnosis method uses some historical and empirical knowledge of the battery to design some diagnostic rules for fault diagnosis . The model-based approach is to establish a physical model of the battery, which is generally capable of accurately calculating the values of the parameters of the battery.…”