Empirical hysteresis

[rtimms]

We currently support an empirical “current sigmoid” model for OCV hysteresis, where users can give a lithiation and delithiation OCV and switch between them depending on the sign of the current. Including such a model has knock-on effects for other model parameters, which may also need to be depend on the sign of the current in order to accurately fit experimental data. In particular, we should add empirical hysteresis for exchange-current and diffusivity.

Longer term we plan to add more sophisticated and self-consistent hysteresis models, but this is a good pragmatic solution for now.