Short circuit fault detection and quantitative analysis based on mean-difference model with VMD

Abstract

Short circuit failure is one of the triggers for thermal runaway of lithium-ion batteries, which can lead to serious safety issues. This paper attempts to estimate the short-circuit resistance of the cell using the mean difference model and relies on the estimated results to make a quantitative analysis of short-circuit fault. To achieve this goal, a combination of forgetting factor recursive least squares and extended kalman filter is used to estimate the average open-circuit voltage within the battery pack. Subsequently, since both the open-circuit voltage (OCV) and intrinsic mode function (IMF0) components reflect the low-frequency characteristics of battery voltage, we propose a new method based on the variational modal decomposition to extract the differential open-circuit voltage of the battery, and finally make an estimate of the short-circuit resistance after obtaining OCV of the battery using the idea of the mean difference model (MDM). In addition, the effectiveness of the proposed method is verified under different degrees of short-circuit faults by connecting different resistors to the series battery pack.