Introduces the Kalman filter as a method that can solve problems related to estimating the hidden internal state of a dynamic system. Develops the background theoretical topics in state-space models and stochastic systems. Presents the steps of the linear Kalman filter and shows how to implement these steps in Octave code and how to evaluate the filter’s output.
Overview
Syllabus
- What is the purpose of a Kalman filter?
- This week, you will learn what a Kalman filter is and generally what it does. You will be introduced to the roadmap for the course and the specialization, and will learn some applications that use Kalman filters.
- What do I need to know about state-space models?
- Kalman filters estimate the "state" of a system that is described using a "state-space model." This week, you will learn the background concepts in state-space models that are required in order to implement a Kalman filter.
- What do I need to know about random variables?
- Systems whose state we would like to estimate are affected by unknown inputs ("disturbances" or "process noises") and their measurements are affected by sensor noises. These noises are modeled by random variables. This week, you will learn the background concepts in random variables that are required in order to implement a Kalman filter.
- State-estimation application of a Kalman filter
- Even though we have not yet derived the steps of the Kalman filter, it is instructive to gain insight into a Kalman filter's operation by watching it run. This week, you will learn how to implement a Kalman filter in Octave and see cases where it works well and where it fails (next course, you will learn why!).
Taught by
Gregory Plett
