Kalman filter
The Kalman filter is an algorithm (a step-by-step process) that helps people remove errors from numbers. It is named for Rudolf E. Kálmán, a mathematician who helped to make it.
Science can use the Kalman filter in many ways. One important use is steering airplanes and space ships.[1] People also use the Kalman filter to make a model of how humans use nerves and muscles to move their bodies.[2]
The Kalman filter has two steps. The first step is predicting (trying to say what you think will happen). The Kalman filter makes a first guess about what we think is true (an estimate) and how certain we are that it is true (uncertainty). Next, the Kalman filter makes a new guess by using a weighted average. More certain numbers are more important in this weighted average. After doing these two steps, we use the new guess to start these steps again.
Kalman Filter Media
Model underlying the Kalman filter. Squares represent matrices. Ellipses represent multivariate normal distributions (with the mean and covariance matrix enclosed). Unenclosed values are vectors. For the simple case, the various matrices are constant with time, and thus the subscripts are not used, but Kalman filtering allows any of them to change each time step.
Truth; filtered process; observations.
A Kalman filter as a Hidden Markov model .
References
- ↑ Paul Zarchan; Howard Musoff (2000). Fundamentals of Kalman Filtering: A Practical Approach. American Institute of Aeronautics and Astronautics, Incorporated. ISBN 978-1-56347-455-2.
- ↑ Wolpert, Daniel; Ghahramani, Zoubin (2000). "Computational principles of movement neuroscience". Nature Neuroscience. 3: 1212–7. doi:10.1038/81497. PMID 11127840. S2CID 736756.
