If you want to know why engineers use Kalman filters and why this is such a popular technique for state estimation, watch this video first before moving on to the next ones below. If you instead want to dive right into the workings of Kalman filters, you can skip this video.

A Kalman filter is an optimal state observer. Before we get to the optimality of Kalman filters, first gain an understanding of how state observers work and discover the math behind them.

Initially, it might be difficult to grasp Kalman filter concepts if you’re starting directly with the state estimation algorithm. Instead, walk through this example which explains visually the working principle of Kalman filters.

Once you have a basic understanding of Kalman filters, you can watch this video to learn how Kalman filter algorithm computes the equations associated with the prediction and update steps.

A Kalman filter is defined for linear systems only so how do we deal with state estimation of real life systems as they mostly exhibit nonlinear behavior? There are different variations of Kalman filters such as Extended, Unscented Kalman filters and Particle filters that you can use for state estimation of nonlinear systems.

If you're not familiar with Simulink, this is a great resource to learn the basics of creating and simulating models in Simulink. Otherwise, feel free to proceed to the next step to practice design and implementation of Kalman filters.

Try it yourself! Download the Kalman filter virtual lab and practice design of a linear and extended Kalman filters using a simple pendulum example. You’ll be able to test your design with simulation and pendulum animations.