Maximum Allowable Current Prediction for Battery Systems Considering Intra-Horizon Parameter Update

With the rapid development of transportation electrification as well as grid-level energy storage, large-scale battery systems have been vastly deployed, and their current or power needs to be well restricted for the safe and efficient operation. While the operation limits for battery cells are commonly specified by manufacturers, the real-time maximum allowable charge and discharge currents of an entire battery system composed of series- or parallel-connected cells/modules/packs, also referred to as the battery system's state of power (SoP), need to be predicted on the fly. To do so, the first challenge is how to deal with the complex dynamic current distribution among parallel-connected battery cells or modules. Moreover, for the long SoP prediction horizon, some battery parameter values vary with the battery state evolution. To address these challenges, this paper proposes an accurate SoP prediction method for parallel-connected battery systems considering parameter update within the prediction horizon. The prediction precision is evaluated by simulation tests based on battery modeling parameter values extracted from charge and discharge experiments of commercial lithium-ion batteries.