Learning Kalman filter basics through simulation

This paper demonstrates how characteristics of Kalman filtering can be illustrated using simulation exercises. The simulation exercises consist of estimating piecewise constant parameters and estimating the state vector of a dynamic system in both the open-loop and the closed-loop setting. Through the use of these simulation exercises, strong mental images of Kalman filtering concepts are quickly captured. Some of the concepts that are demonstrated are covariance resetting, utilizing a constant Kalman gain matrix, the importance of accurately modelling the system dynamics, and the effectiveness of Kalman filtering in reducing the unwanted system perturbations caused by measurement noise and system noise.