Graphics Programs Reference
In-Depth Information
velocity for the
(
n
+
1
) th
observation. An implementation of the
αβγ
tracker
is shown in Fig. 9.21 .
The tracker will follow an input whose acceleration is constant with no
steady state errors. Again, in order to reduce the error at the output of the
tracker, a weighted difference between the measured and predicted values is
used in estimating the smoothed position, velocity, and acceleration as follows:
αβγ
x s
() x p
=
() α x 0
+
(
() x p
–
()
)
(9.110)
β
---
ß s
() ß s
T ßß s
=
(
n
–
1
)
+
(
n
–
1
)
+
(
x 0
() x p
–
()
)
(9.111)
T 2
ßß s
() ßß s
-----
=
(
n
–
1
)
+
(
x 0
() x p
–
()
)
(9.112)
() T 2
2
() T ß s
ßß s
-----
x p
(
n
+
1
)
=
x s
+
+
()
(9.113)
and the initialization process is
x s
() x p
=
() x 0
=
()
ß s
() ßß s
() ßß s
=
=
() 0
=
x 0
() x 0
–
T
()
ß s ()
=
--------------------------------
-
+
+
x p
x 0
+
Σ
Σ
Σ
α
–
1
delay, z
+
+
+
x s
T
ß s
ß p
β
---
+
+
Σ
–
1
Σ
delay, z
+
+
ßß s
T 2
T
2
T 2
-----
Σ
–
1
delay, z
+
+
αβγ
Figure 9.21. An implementation for an
tracker.
 
Search WWH ::




Custom Search