Advanced Battery Management Algorithms
My thesis work was devoted to developing advanced finite element models and adaptive observers of batteries; almost a gas gauge, but for a battery. When a piece of tech, like your phone or laptop, is using battery power, it only trickles a small amount of energy from the battery at a given time. This means that conventional battery management systems, operating in linear regimes, can handle predicting the current state of charge under those conditions. When something like an electric vehicle is drawing charge from a battery, though, it varies wildly in magnitude of power consumed. That's why we need better models, that more accurately recreate those high discharge/charge scenarios, and observers to wrap around them, to better predict the time-dependent change of battery parameters. This graph depicts an observer estimating the physical parameters of a battery (normalized to one, so they can all fit on the same y-axis), using a novel Jacobian adaptive observer. Lots of cool math went into this, which I'm always into.