Unscented Kalman Filter Project Starter Code

Workflow

The workflow is the same as the EKF fusion project, but there are two differences:

• UKF used instead of EKF to fusion Radar and Laser data.
• ctrv model used as the process model with 5 parametes in the state vector.

Input data format

Input data is located in the "data" directory. Here is the data format description.

#Input file format:
#L(for laser) meas_px meas_py timestamp gt_px gt_py gt_vx gt_vy
#R(for radar) meas_rho meas_phi meas_rho_dot timestamp gt_px gt_py gt_vx gt_vy
-----------------------------
Example
-----------------------------
R	8.60363	0.0290616	-2.99903	1477010443399637	8.6	0.25	-3.00029	0
L	8.45	0.25	1477010443349642	8.45	0.25	-3.00027	0 

Results

We have generated two simulated tracking data to test our implementation, here is the tracking results. We noticed that the UKF fusion resutls is more smooth than EKF at the circle part, and we archive a smaller RMSE values.

We also found that it is difficult to archive the best RMSE and the best filter consistancy (indicated by the NIS figures), by only adjusting the process noise parametes.In this project, we scrifice the NIS to archive a little bit to archive best RMSE values.

• Track 1. px, py, vx, and vy RMSE: [0.0740189, 0.0837208, 0.583002, 0.586191] meters

• Track 1 Radar NIS

• Track 1 Laser NIS

• Track 2. px, py, vx, and vy RMSE: [0.197399, 0.189804, 0.548703, 0.543699] meters

• Track 2 Radar NIS

• Track 2 Laser NIS

Dependencies

• cmake >= v3.5
• make >= v4.1
• gcc/g++ >= v5.4

Basic Build Instructions

1. Clone this repo.
2. Make a build directory: mkdir build && cd build
3. Compile: cmake .. && make
4. Run it: ./UnscentedKF path/to/input.txt path/to/output.txt. You can find some sample inputs in 'data/'.
   • eg. ./UnscentedKF ../data/sample-laser-radar-measurement-data-1.txt output.txt