Global Positioning System Reference
In-Depth Information
The design matrix
H
is:
100000000000000
010000000000000
001000000000000
GPS
k
H
=
(32)
000100000000000
000010000000000
000001000000000
v GPS
And
k is a white noise process with zero mean and unity variance. When GPS is not
available, forward velocity derived from the wheel encoders together with pitch and azimuth
estimates are used as a measurement update. The measurement model is as follows:
z odo
k
odo
k
v odo
k
= H
x k +
(33)
z odo
k
Where the measurement state vector
is defined as:
V INS
e k
V odo
e k
V INS
n k
V odo
n k
z odo
k
=
(34)
V INS
u k
V odo
u k
H
The design matrix
is:
000100000000000
000010000000000
odo
k
H
=
(35)
000001000000000
v odo
k
And
is a white noise process having zero mean and unity variance.
3. Experimental setup
3.1Wheeled mobile robot
Outdoor trajectory tests are conducted to assess the performance of the developed navigation
solution; the tests are conducted using a mobile robot shown in fig 2. The robot was developed
by a member of NavINST and members of the Electrical and Computer Engineering
Department (ECE) at Royal Military College (RMC) of Canada. The mobile robot is
three-wheeled and differentially-driven with a quadrature optical encoder coupled to the
drive outputs of each motor. Appropriate scaling in the navigation scheme is used to provide
angular velocity estimates of each wheel from these encoders. Power on-board the mobile
robot comes from three sources, namely (1) two 12V batteries connected to the drive amplifiers
and motors, (2) two 12V batteries connected to the sensors and encoder processor board and
(3) a battery internal to the laptop mounted on the top level of the robot.
Search WWH ::




Custom Search