Virtual AnnaAI

The work in brief

Anna's work applies control-oriented modeling and estimation to lithium-ion batteries, aiming to infer internal states that cannot be measured directly from limited sensing. Early papers (2006–2012) address battery sizing and power split in fuel cell hybrid vehicles, electrochemical-model-based Kalman filters for state of charge and critical surface concentration, and neutron imaging of lithium concentration to validate such estimators. Through the 2010s the work broadens to lumped electro-thermal models and core-temperature observers for cylindrical and prismatic cells, sparse-sensor estimation in series strings, and energy-optimal warm-up from subzero temperatures. A distinctive strand uses mechanical signals: phenomenological force and swelling models, force-based incremental capacity analysis, and expansion measurements added to voltage to improve state-of-charge and electrode-specific state-of-health (eSOH) estimation.

More recent papers (roughly 2019–2026) move toward degradation, safety and lifecycle questions. They include eSOH identifiability and error bounds with partial data, accelerated lifetime simulation, data-driven and domain-informed lifetime prediction, and a "deepSOH" argument that path-dependent degradation requires full state estimation aided by expansion sensing. Safety work covers internal and external short circuits, venting and thermal runaway modeling, carbon dioxide gas sensing, and emergency discharge under temperature and pressure constraints. The group also studies manufacturing formation (SEI growth models, end-of-line differential voltage diagnostics) and the analysis of parallel-connected cell imbalance, retired-pack screening, second-use storage power electronics, and field data from grid storage systems.

AI-written from the paper titles and abstracts; Anna hasn't reviewed it.