The team consisting of...
The Kalman filter is a set of mathematical equations that provides an
efficient computational estimate of the state
of a process (e.g., the position and orientation of an airplane)
given a
time-varying sequence of noisy
measurements (e.g., air speed, pressure, temperature, engine thrust).
The filter is a popular mathematical estimator due to its efficiency
and robustness.
Our goal is to develop a web-based tool to help
develop the intuition and insight of
novice users regarding the behavior of the Kalman filter.
Users would have the ability to change various input parameters
and then see how the Kalman
filter responds for a given set of noisy measurements.
Because the software is to be primarily used as a teaching aid, we
have several design goals.
First,
the interface should be simple, elegant, and easy to use.
Input choices will be limited as necessary to avoid
overwhelming the user.
Parameters will be restricted to realistic but interesting values.
A single,
concrete physical example
(e.g., sensing water level in a tank)
will be used as well for all
simulations.
Performing multiple simulations using different physical models can
lead to confusion for a novice user.
Second,
the output from a simulation should be organized and displayed in a
manner that is easy to understand.
Users should be able to see a (graphical) overview of the filter's
estimation of the process state and how it compares to the actual
state.
Users should also be able to step through the measurement data and the
equations that comprise the Kalman filter and trace the intermediate
steps of the algorithm.
This helps in understanding the inner workings of the Kalman filter.
Finally, the resulting software should be easy to extend, modify, and
maintain.
The design of the software should not preclude for example
the use of different physical examples or the addition of input
choices and parameters.
Use The Kalman Filter On-line Learning Tool
Library of Documents