College of Engineering

ECE MS Thesis Defense by John K. Chabior

Date(s): 12/14/2011 1:30 PM - 12/14/20113:30 PM
Location: Lester W. Cory Conference Room, SENG 213A
Contact: Steven C Nardone snardone@umassd.edu 508-999-8510


TOPIC:“TECHNIQUE FOR 3-DIMENSIONAL DIGITAL MAGNETIC COMPASS CALIBRATION”

ABSTRACT:
Explorers and navigators have used the Earth’s magnetic field to find their direction for centuries. Even with the introduction of GPS technology, many high precision navigation systems employ some sort of magnetic compass as a back-up reference. Magnetically distorting materials often surround these compasses, introducing significant heading error. Therefore, it is important to calibrate the compass sensors to compensate for these distortions. This thesis presents theoretical, simulation, and real experimental results of a 3-dimensional calibration algorithm developed for an electronic magnetic compass.

Additionally, strap down compassing systems have become increasingly popular as they are much less mechanically complex compared to the gimbaled system. One difficulty with using strap down systems is that the scheme is much more computationally complex, but as transistor and processing technology continues to improve this poses less of a problem. Another obstacle in providing accurate measurements with the strap down system is filtering the vehicle’s vibrations from the attitude measurements, as this can introduce a significant error in attitude and heading readings. This thesis shows the development and results of a 2-state Kalman filter used to dampen mechanical vibrations of the sensors.

Results with synthetic data are very promising for both the Kalman filtered accelerometer data and magnetic field calibration. Filtered data which simulates shaking of the accelerometers is dampened significantly while minimizing overshoot and time-delay. Unfortunately, the gyroscopes on the actual system do not perform as expected and as a result real data is not filtered as well as synthetic data. In order to resolve this problem a process noise adaptive Kalman filter is developed. Similarly, synthetic shapes mimicking the distorted magnetic field are calibrated well and results are very good. Results with real distorted measurements taken by the compass look promising but are not as accurate as with synthetic data. The experiments show that heading error can be minimized to within a few degrees using this calibration scheme. The general concept of the Earth’s magnetic field, navigation, magnetic distortions, and the sensor system as a whole is discussed in this thesis along with the development of the filtering and calibration routines.

NOTE: All ECE Graduate Students are ENCOURAGED to attend.
All interested parties are invited to attend.
Open to the public.

Advisor: Dr. Steven C. Nardone
Committee Members: Dr. Dayalan Kasilingam and Mr. John S. Davis


Contact Info: