A C++ implementation of the Unscented Kalman Filter (UKF) using Eigen. The UKF uses a deterministic sampling technique known as the Unscented Transformation to calculate statistics around the mean. This technique does not require differentiability of the models.
See also https://groups.seas.harvard.edu/courses/cs281/papers/unscented.pdf for more details.
For the implementation I took inspiration from https://github.com/CoffeeKumazaki/kalman_filters.
Tested only on Ubuntu 20.04 LTS
- Eigen3 library.
Clone the repo on your PC. Then, enter the folder where you downloaded the repo, open a terminal and run:
mkdir build && cd build
cmake .. -DCMAKE_INSTALL_PREFIX="/path/to/desired/install/dir"
make install
The UKF library will be available in the install/lib folder. The library is composed of two classes:
-
UnscentedKF: implementation of the Unscented Kalman Filter in c++. Provides methods to set up a routine for estimation/filtering of user-defined quantities via UKF. Works alongside with the UKFModel class which provides the prediction and observation models.
-
UKFModel: defines the prediction and observation models for the UnscentedKF class. This is an abstract class: the user must create his own child class derived from UKFModel, and implement with it the prediction and observation models for his specific problem.
An example of usage of the library is available in the example folder. The filter is used to estimate the thrust provided by a jet engine and its rate of change, given the measured thrust data.
A simple test to verify library integrity is provided in the test folder. The test uses Catch2 library and can be run with the ctest command from the build directory.
Gabriele Nava, @gabrielenava