Implementation in MATLAB of a Multiplicative Extended Kalman Filter for live estimation of a smart device's attitude

Abstract

This work introduces the tools used to teach the Kalman Filter (KF) to Aerospace Engineering students in the University of Seville. In particular, an easy-to-set-up application is introduced; based on the Matlab framework on its 2020b (or newer versions), it is able to display the attitude of a smart device in real time through a wireless connection to a computer. This tool is a simple yet powerful educative resource when teaching about the KF, since it showcases its performance while allowing the student to understand how a complex real-world problem can be solved using a relatively simple implementation of the KF; in particular, the Multiplicative Extended KF (MEKF) is chosen, but the framework can be easily adapted to other versions such as the Extended or Unscented KF. In addition, the tool allows the student to be aware of the inner workings of the filter itself, learning about its advantages and limitations compared with other attitude estimation algorithms. The student is also able to understand how the filter needs to be tuned, and to observe the results of the experiments in a visual and straightforward manner.