Title: Development and Testing of a Real-Time GPS/INS Reference System for Autonomous Automobile Navigation
Abstract: Precise position and velocity information have already led
to a large number of vehicle-based applications. Access
to this information in real-time further opens the door to
telematic applications such as en-route navigation
assistance. This paper deals with the development of a
real-time system to be used for the testing of proposed
autonomous automobile navigation systems, whereby the
driver need not actually control the vehicle. From a
navigation standpoint, this requires not only very precise
but also very reliable position and velocity information
with high availability and low latency. These constraints
are even more stringent for a reference system, whose
solution is to be considered ideal, as is the case here. To
this end unaided GPS is insufficient, as it does not
provide the necessary availability or reliability. A loosely
integrated approach using a GPS Real-Time Kinematic
(RTK) and mid-level INS is therefore proposed to
alleviate these problems.
Given the time-critical nature of such an application,
emphasis is given to the real-time aspect of the system.
An implementation is presented whereby IMU data are
processed as they are received and INS states are updated
with inherently latent GPS data. The objective of this
paper is to assess the accuracy and feasibility of such a
system, which is based on the Honeywell HG-1700 IMU
and the NovAtel OEM4 GPS receivers.
A vehicle test was performed under varying operational
conditions, including open sky, and areas where shading
is encountered due to buildings, foliage and overpasses.
System performance is evaluated through the accuracy of
the INS only solution during periods of simulated GPS
outages. Initial testing shows the system is able to bridge
short GPS data gaps (1 to 5 seconds) with decimeter level
accuracy under mild vehicle dynamics. Increased vehicle
dynamics combined with longer GPS outages currently
produce INS errors at the meter level.
Publication Year: 2001
Publication Date: 2001-09-14
Language: en
Type: article
Access and Citation
Cited By Count: 18
AI Researcher Chatbot
Get quick answers to your questions about the article from our AI researcher chatbot