Difference between Extended Kalman Filter and Augmented Extended Kalman Filter

I have taken up basic courses on MATLAB and Simulink knowing the concepts and developed some basic models and gained knowledge, but still i am new to many things. Now , to learn more ,I am developing a Simulink battery model to estimate state of health of a battery using MATLAB/Simulink. I can use Unscented Kalman Filter but rather I want to know about Augmented Extended Kalman Filter.

Also the basic idea to simulate state of health of battery by using simulink model is by estimating the growth in its Internal Ohmic Resistance with Increase in temperature which results in aging process of battery (both Cycle aging and Calender aging). So the EKF or AEKF will perform online parameter estimation to determine this Internal Resistance. So based on priori estimates ,how come exactly this internal resistance given by online parameter estimation , what exactly happens in recursive algorithm in this case and how to define state and maesurement update function in m-script for battery, i need bit clarification.

Answer 1

[Joel Van Sickel]

Hello Sanya Gode,

here is a document listing the difference between our unscented and extended kalman filters: https://www.mathworks.com/help/control/ug/extended-and-unscented-kalman-filter-algorithms-for-online-state-estimation.html

if you aren't aware, here is an example using kalman filters for battery control: https://www.mathworks.com/help/control/examples/nonlinear-state-estimation-of-a-degrading-battery-system.html

As to the best choice of filter, that will depend on your particular needs and implementation. Hopefully this helps you design some experiments to learn more about the trade space of different filter implementaitons.

Regards,

Joel