UWE Bristol Engineering showcase 2015

Paven Bhatti
MENG Aerospace Engineering
Development and Implementation of Magnetometer Drift Correction
in an IMU Navigation System
Project Brief
This MENG project focuses on the implementation of a
yaw drift correction method for a standard IMU system.
The results from any testing will be critically analyzed to
determine whether the system would be suitable for
implementation into a UAV fly-by-wire system. This is the
second of a two-phase project designed to complete an
IMU system capable of being tested in a scale glider. If
this project were successful it would create the potential
for creating a completely independent fly by wire
system, capable of long-term flight, as well as readily
available use in GPS dead signal zones such as
mountainous terrain.
Project Aims and Objectives
The aim of this second phase of a two-phase project is to complete
the goal of creating a completely independent IMU system, with
the potential of implementation into a fly by wire system to create
an independent navigational system capable of long-term flight. It is
aimed to understand and implement an internal yaw drift
correction method onto a standard IMU. Following this it is aimed
to test the successful implementation and collect data that can be
analyzed from an aircraft’s navigational perspective, and following
that attempt to implement the finished IMU system into a basic
servo responsive code that would react to the change in pitch, roll
and yaw. It is also aimed to continue the DO-178 B as well as CAA
safety requirements outlined in phase one of the project.
Phase 2’s objectives are detailed below
Assemble and run the chosen IMU configuration with yaw drift
corrective method
Successfully implement a code to react to changes in the sensory
data from the IMU
Use the created code to fly the UAV glider to a predetermined
location, with the optional aid of the controller design chosen
Introduction
Previously in this project research was undertaken in order to understand
the concept of creating a fly by wire system capable of navigating a UAV,
or similar craft through a predetermined set of co-ordinates in space. The
previous results showed that without the aid of a GPS system to correct
the “drift” in the yaw axis the IMU was incapable of determining its own
position in space accurately for longer than a few seconds before the
margin of error was large enough to render its data output invalid.
In the second phase of this project, suitable yaw drift correction methods
will be explored. The traditional GPS drift correction method will be
compared to the use of magnetometers, which will be used in an attempt
to placate the need for a GPS device. Magnetometers may be able to
accommodate the IMU’s need for drift correction as it is capable of using
the geomagnetic field as a constant vector with which to gather a
reference point. It is hypothesized that knowledge of the earth’s
magnetic field-patterns may be able to be used to correct the drift
obtained by the IMU’s ability to determine its position in space during
motion. Assuming a magnetometer can sufficiently replicate the GPS’s
effects, the functioning stabilization system will then be implemented
into a UAV navigational code to test and demonstrate the success of the
system, in order to gain proper data with which to compare the two
methods of yaw drift correction.
Project Supervisor
Dr. Pritesh Narayan
Project Conclusion
In conclusion the majority of the aims and
objectives of this phase of the project were
fulfilled. The IMU now has an independent
method for obtaining yaw drift correction,
and that method is accurate enough to be
tested in a UAV or similar test bed in order to
assess its ability to lead an active-servo
control system.
There were drawbacks with the
implementation of the Kalman filter into an
IMU code, which resulted in the requirement
to use an external magnetometer instead of
one that was available and working in the
MPU9150.