2002 Florida Conference on Recent Advances in Robotics
1
Inertial Navigation
Kevin J Walchko
1
University of Florida, Gainesville, FL 32611-6200
Dr. Paul A. C. Mason
2
NASA Goddard Space Flight Center, Greenbelt, MD
This paper will discuss the design and implementation of an inertial navigation system (INS) using an inertial
measurement unit (IMU) and GPS. The INS is capable of providing continuous estimates of a vehicle’s position
and orientation. Typically IMU’s are very expensive sensors, however this INS will use a “low cost” version
costing only $5,000. Unfortunately with low cost also comes low performance and is the main reason for the
inclusion of GPS into the system. Thus the IMU will use accelerometers and gyros to interpolate between the
1Hz GPS positions. All important equations regarding navigation are presented along with discussion. Results
are presented to show the merit of the work and highlight various aspects of the INS.
I. Introduction
Navigation has been present for thousands of years in some
form or another. The birds, the bees, and almost everything
else in nature must be able to navigate from one point in
space to another. For people, navigation had originally
included using the sun and stars. Over the years we have
been able to develop better and more accurate sensors to
compensate for our limited range of senses. This paper will
discuss work using one of these advanced sensors, an
inertial measurement unit (IMU). This sensor, coupled with
the proper mathematical background, is capable of
detecting accelerations and angular velocities and then
transforming those into the current position and orientation
of the system.
Inertial Navigation Systems (INS) have been developed for
a wide range of vehicles. Sukkarieh [1] developed a GPS/
INS system for straddle carriers that load and unload cargo
ships in harbors. When the carriers would move from ship
to ship, they would periodically pass under obstructions
that would obscure the GPS signal. Also, as the carriers got
closer to the quay cranes, it became more difficult to get
accurate positions due to the GPS signal being reflected
about the cranes metal structure. This increases the time of
flight of the GPS signal and results in jumps in the position.
During these times the INS would then take over, and guide
the slow moving carrier until a reliable GPS signal could be
acquired.
Bennamoun et al [2] developed a GPS/INS/SONAR system
for an autonomous submarine. The SONAR added another
measurement to help with accuracy, and provided a
positional reference when the GPS antenna got submerged
and could not receive a signal.
Ohlmeyer et al [3] developed a GPS/INS system for a new
smart munitions, the EX-171. Due to the high speed of the
missile, update rates of 1 second from a GPS only solution
were too slow, and could not provide the accuracy needed.
A. Outline
The first section of this paper will introduce inertial
navigation. Then the IMU and GPS hardware will be
covered. Finally experimental results using this INS will be
presented.
II. Inertial Navigation
This section will cover strap-down inertial navigation by
first describing the methods and equations. Next sources of
error for these systems and how the kalman filter will be
utilized to account for these errors.
A. Overview of Inertial Navigation Systems
A basic flow chart of how inertial navigation works is
shown in Figure 1. However, this is not all that needs to be
done to have an INS that works. There are many problems
with noise and unbounded error that must be handled to get
any meaningful result out of the INS.
Gimballed INS
The first type of INS developed was a gimballed system.
The accelerometers are mounted on a motorized gimballed
platform which was always kept aligned with the
navigation frame. Pickups are located on the outer and
inner gimbals which keep track of the attitude of the
stabilized platform relative to the vehicle on which the INS
is mounted. This setup has several detractors which make it
undesirable.
1. Graduate Research Assistant, Mechanical Engineering. Student member AIAA.
2. Mechanical Engineer, Flight Dynamics Analysis Branch. Member AIAA.