This paper performs a deep Performance Analysis of the most utilised filters for prediction estimation on tracking systems which are the Extended Kalman Filter (EKF) and the Unscented Kalman filter (UKF). A real-time simulation is used to compare both filters and praise the advantages of each approach, where both are compared with their respective MATLAB function. All the performed scenarios aim to test the capabilities of the filters in a real-life scenario and, in the end, discover what are the best conditions to use each method.
Report: file
Papers Grade: 17/20