Code files

kf.zip
63 Kb

All code necessary to run the Kalman Filter. Includes:

  • kf.m - the main script. With all these files in the same directory, you can run the kalman filter by running kf.
  • show.m - script to plot the ground truth and localized paths, along with the beacon positions, after running kf.m.
  • showDR.m - same as show, but also plots the path from dead reckoning.
  • errors.m - script to compute several error statistics about the localized path. See below for a description of the error statistics.
  • several other functions called by these scripts.

 

Localization Results

The error stats considered here are:

  • Cross-Track Error (XTE) -- the component of error orthogonal to the true robot path
  • Along-Track Error (ATE) -- the component of error tangential to the path

The errors.m script also computes

  • Cartesian error, based on the Cartesian distance between the true position and the estimate at each point in time.

Data Set A1

  

Error Statistics:

 

XTE

ATE

Mean Absolute

0.5927m

0.4905m

Maximum

2.1526m

2.2027m

Standard Deviation

0.4137m

0.3686m

Data Set A2

  

Error Statistics:

 

XTE

ATE

Mean Absolute

0.6077m

0.5433m

Maximum

1.7471m

1.6905m

Standard Deviation

0.3996m

0.3653m

Data Set B1

  

Error Statistics:

 

XTE

ATE

Mean Absolute

4.8993m

5.5533m

Maximum

14.3326m

15.8426m

Standard Deviation

2.9577m

3.7952m

Data Set B2

  

Error Statistics:

 

XTE

ATE

Mean Absolute

0.6880m

0.5885m

Maximum

2.6127m

2.9487m

Standard Deviation

0.4968m

0.4833m

Data Set C1

  

Error Statistics:

 

XTE

ATE

Mean Absolute

0.5539m

0.3976m

Maximum

1.9033m

2.0447m

Standard Deviation

0.4173m

0.3558m

Data Set C2

  

Error Statistics:

 

XTE

ATE

Mean Absolute

0.2827m

0.2568m

Maximum

0.7672m

0.8391m

Standard Deviation

0.1717m

0.1975m

Data Set C3

  

Error Statistics:

 

XTE

ATE

Mean Absolute

0.2726m

0.3249m

Maximum

1.4054m

2.9122m

Standard Deviation

0.2473m

0.2788m