Target Tracking - MATLAB Simulations for Radar Systems Design - page 449

9.10. ÐMyRadarÑ Design Case Study - Visit 9

9.10.1.Problem Statement

Implement a Kalman filter tracker into the ÐMyRadarÑ design case study.

9.10.2. A Design 1

For this purpose, the MATLAB GUI workspace entitled Ðkalman_gui.mÑ
was developed. It is shown in Fig. 9.31 . In this design, the inputs can be initial-
ized to correspond to either target type (aircraft and missile). For example,
when you click on the button ÐResetMissile,Ñ the initial x-, y-, and z- detection
coordinates for the missile are loaded into the ÐStarting LocationÑ field. The
corresponding target velocity is also loaded in the Ðvelocity in x directionÑ
field. Finally, all other fields associated with the Kalman filter are also loaded
using default values that are appropriate for this design case study. Note that
the user can alter these entries as appropriate.

This program generates a fictitious trajectory for the selected target type.
This is accomplished using the function Ðmaketraj.mÑ . It is given in Listing
9.6 in Section 9.11. The user can either use this program, or import their own
specific trajectory. The function Ðmaketraj.mÑ assumes constant altitude, and
generates a manuevering trajectory in the x-y plane, as shown in Fig. 9.32 . This
trajectory can be changed using the different fields in the Ðtrajectory Parame-
terÑ fields.

Next the program corrupts the trajectory by adding white Guassian noise to
it. This is accomplished by the function Ðaddnoise.mÑ which is given in List-
ing 9.7 in Section 9.11. A six-state Kalman filter named Ðkalfilt.mÑ is then uti-
lized to perform the tracking task. This function is given in Listing 9.8.

The azimuth, elevation, and range errors are input to the program using their
corresponding fields on the GUI. In this example, these entries are assumed
constant throughout the simulation. In practice, this is not true and these values
will change. They are caluclated by the radar signal processor on a Ðper pro-
cessing intervalÑ basis and then are input into the tracker. For example, the
standard deviation of the error in the range measurement is

∆ R c

σ R = ------------------------ = --------------------------------

(9.161)

2 × SNR 2 B 2 × SNR

1. The MATLAB code in this section was developed by Mr. David Hall, Consultant to
Decibel Research, Inc., Huntsville, Alabama.