Accuracy of shipborne kinematic GPS surveying
Loading...
Authors
Grinker, Barry
Subjects
GPS
Kinematic
Kinematic
Advisors
Clynch, J.R.
Date of Issue
1991-09
Date
Publisher
Monterey, California. Naval Postgraduate School
Language
en_US
Abstract
In December 1990 an experiment was conducted at the Naval Postgraduate School
in Monterey, California in which four different receivers, mounted on the mast of a ship,
collected data simultaneously for several hours at a time. Ashtech LD -XII, Trimble
4000 ST, TI 4100 and Magnavox MX 4200 receivers were used.
The reference system consisted of a Krupp Atlas Polarfix laser system set up on the
shore at a pre-surveyed site. A two-axis vertical gyro system and a heading gyro gave the
ship's 3-dimensional orientation at any instant in time, providing a connection between
the laser reflector and the GPS antennas on the ship's mast. This enabled the reduction
of the laser reflector's trajectory to the Ashtech and Trimble antennas for subsequent
comparison to the kinematic GPS trajectories of these receivers determined by the
postprocessing of the data collected. Each data set was processed once with the software provided by the manufacturer
and once with an independent software package, OMNI, developed by the National
Geodetic Survey. In addition to the software, six factors were examined to determine
their effects on kinematic GPS surveys. They included : tropospheric corrections, initialization,
satellite geometry, ephemeris type, data interval and multipath.
In general the software available is versatile; however each package has some limitations
which call for further development. The results indicate that the effects mentioned
are small, generally less than the noise in the reference system. An important
factor is the initialization of the kinematic process. Bad initialization can cause a reasonable
solution for a period of time, but a sudden deterioration when the satellite configuration
changes. Accuracy levels of a few decimeters were easily achieved with the
systems and procedures used. Both Ashtech and Trimble produced trajectories which
were accurate to within the noise level of the laser trajectories. In both cases the solution
produced by OMNI differed from the solution produced by the manufacturer's software,
only by a few centimeters.
Type
Thesis
Description
Series/Report No
Department
Hydrographic Science
Organization
Naval Postgraduate School
Identifiers
NPS Report Number
Sponsors
Funder
Format
106 p.: ill.;28 cm.
Citation
Distribution Statement
Approved for public release; distribution is unlimited.