Integration of GPS with Other Sensors and Network Assistance - Understanding GPS: Principles and Applications - page 502

pitch and roll has a second-order effect upon the gyro scale factor, as indicated in
(9.14), but this should be small relative to its nominal scale factor error.
Given the preceding discussion on inertial system options, the error characteris-
tics of gyros and accelerometers can now be addressed. For the low-cost sensors con-
sidered for automotive applications, the bias and scale factor errors can be very large
relative to those of gyros and accelerometers associated with navigation grade sys-
tems (e.g., for the gyro, a bias of several degrees per second is expected, and a scale
factor error as large as 5% is possible). A summary of gyro and accelerometer bias
and scale factor errors for different applications may be found in [25]. Of course,
these errors can be calibrated using GPS and other means. For instance, an estimate
of the gyro bias can be obtained each time the vehicle is stationary in a calibration
procedure referred to as a zero velocity update (ZUPT). However, the errors can
also be quite unstable and have high temperature sensitivities.
Figure 9.25 illustrates the laboratory-measured gyro bias temperature sensitivi-
ties for two samples of a low-cost, vibrational gyro. The term vibrational indicates
that the gyro has a vibrating element that senses angular rate through a Coriolis
force exerted on the vibrating element. This force is directly proportional to the
angular rate and is measured through the actions of the gyro electronics. Figure
9.26, abstracted from [29], illustrates the driving and detection and control mecha-
nisms for the Murata Gyrostar gyro as an example of vibrational gyro technology.

100

80

60

40

20

0

− 20

− 40

− 60

− 80

− 40.0 − 20.0 0.0 20.0 40.0 60.0 80.0 100.0

Temperature (ºC)

EM 0 (BIAS)
Linear fit
EM 4 (BIAS)
Sine fit

Figure 9.25 Bias versus temperature for two low-cost vibrational gyro samples.