Target Tracking - MATLAB Simulations for Radar Systems Design - page 433

Note that Ðghk_tracker.mÑ uses MATLABÓs function Ðnormrnd.mÑ to gener-
ate zero mean Gaussian noise, which is part of MATLABÓs Statistics Toolbox.
If this toolbox is not available to the user, then Ðghk_tracker.mÑ function-call
must be modified to

[residual, estimate] = ghk_tracker1 (X0, smoocof, inp, npts, T)

which is also part of Listing 9.2. In this case, noise measurements are either to
be considered unavailable or are part of the position input array.

To illustrate how to use the functions ghk_tracker.m and ghk_tracker1.m,
consider the inputs shown in Figs. 9.22 and 9.23. Fig. 9.22 assumes an input
with lazy maneuvering, while Fig. 9.23 assumes an aggressive maneuvering
case. For this purpose, the program called Ðfig9_21.mÑ was written. It is given
in Listing 9.3 in Section 9.11.

Figs. 9.24 and 9.25 s how the residual error and predicted position corre-
sponding (generated using the program Ðfig9_21.mÑ ) to Fig. 9.22 for two
cases: heavy smoothing and little smoothing with and without noise. The noise
is white Gaussian with zero mean and variance of . Figs. 9. 26 and
9.27 show the residual error and predicted position corresponding (generated
using the program Ðfig9_20.mÑ) to Fig. 9.23 with and without noise.

σ 2

= 0.05

2.5

2

1.5

1

500 1000 1500 2000 2500 3000 3500 4000 4500 5000

Sam ple num ber

Figure 9.22. Position (truth-data); lazy maneuvering.