Abnormal Overheating Detectability Analysis Based on Cross Gramian for a Supercapacitors String

摘要

Model-based temperature estimation has been widely studied and applied in many applications, including for power/energy storage applications such as batteries or supercapacitors (SCs). In this paper, sensors placement for a temperature monitoring and diagnostic system is studied, with respect not only to the state estimation during normal condition but also abnormal overheating detection. The abnormal overheating is modelled as an unknown additional disturbance. Several measures based on the cross Gramian matrix are used as the basis for the optimal sensors placement for the overheating detectability, and the placement selection is evaluated from different overheating simulation scenarios. A "naive" sensor placement which intuitively places the sensors in an evenly distributed manner along the string is also included to confirm the usefulness of the cross Gramian measure-based sensors placement.