Problem Statement: Battery modules composed of series–parallel cell assemblies exhibit non-uniform ageing due to current imbalance, thermal gradients, and heterogeneous operating histories. Conventional models used in battery management systems inadequately capture coupled electrochemical–thermal degradation across cells under realistic, variable charge–discharge cycles. This limits accurate life prediction, safe operation, and degradation-aware control at the module level.

Approach: A digital twin of a lithium-ion battery module will be developed using two complementary paradigms. A physics-based layer will integrate reduced-order electrochemical or equivalent-circuit models with thermal dynamics and semi-empirical ageing mechanisms (capacity fade, resistance growth). In parallel, AI/ML models (e.g., LSTM or physics-informed neural networks) will learn degradation patterns from cycling data and adaptively correct model uncertainties.

Expected Outcome: The project will deliver a validated module-level digital twin capable of simulating performance and degradation under diverse operating profiles. The framework will enable cycle-life prediction, cell imbalance analysis, and degradation-aware operational strategies, supporting advanced battery management and energy system optimisation.