What does Kalman filter mean?
An estimator that repeats two steps, predict with a model and correct towards a measurement, and works out how far to move towards each measurement from how uncertain the prediction and the measurement are. For a linear system with Gaussian noise no estimator does better.
In a lesson
That is it. The Kalman filter, the particle filter in U7, the whole family: they differ in how the state is represented and how the weighting is chosen, never in the shape.
Learn it
- Taught in U6.2 Predict and correct State estimation, University
- Guide with live demos: Kalman filters explained
Used in
- U3.5 A fix from a landmark
- U4.3 The low pass filter
- U4.4 Outliers
- U6.3 The Kalman gain
- U6.4 Q and R
- U6.5 Covariance and the ellipse
- U7.1 Where am I?
- U7.3 Moving the cloud
- U8.1 A map of cells