Multiscale-multidomain model order reduction of Lithium-ion batteries for automobile application: A review

Lithium-ion batteries have become a preferred choice for energy storage because of having high energy density, lower discharge rate, and long cycle life than other battery chemistries. Due to the widespread use of lithium-ion batteries in electric vehicle power modules, a spike in demand is observed in recent years. Thermal modeling is an efficient way to predict battery thermal behavior. A complete battery model is comprised by an electrochemical model and a thermal model dynamically coupled to simulate battery responses. To improve the design and operation of lithium-ion batteries, multiscale and multi-dimensional modeling approach is used. Due to the involvement of different dimensions and length scales, lithium-ion battery models are complex in nature. To reduce model complexity different approaches are reported in the literature. These approaches are based on model order reduction on different scales i.e., nano-, micro-, and milli-meter scales. Model order reduction is aimed at reducing model complexity while retaining the accuracy. A comprehensive review of different model order reduction approaches is presented. It is found that most of the reduction approaches proposed in published papers focus on reduction on individual levels or scales. A multi-scale multi-domain model order reduction strategy is proposed in the end that involves simultaneous reduction on multiple scales followed by coupling between scales.