In this work, the feasibility of a novel method for combining inertial information gathered from accelerometers to calculate an estimate for rotational information independent of rate gyro measurements is investigated. The concept uses two, distinctly placed, parallel accelerometers on a rigid body to estimate rotational acceleration by comparing the difference in sensed acceleration caused by rotation. Since the estimation is independent of available rate gyro data, optimization of drift and scale factors can be used to calibrate the accelerometers in a real time fashion. Experimental trials were conducted on a small mobile robotic platform, due to the eventual objective of using the kinematical hypothesis described above to develop an improved, compact Inertial Navigation System, ideal for autonomous robotic navigation in hazardous environments where Global Positioning signal are not available.
Preliminary trials were conducted in a two-dimensional environment, and accelerometer data was used to estimate heading and generate positional maps of the paths traveled. The estimated rotation data was also used to roughly determine the scale and drift factors of the accelerometer inputs, which were independently measured for verification. Important conclusions about the effects of instrument placement geometry were reached, and the developed system was scrutinized for recommended improvements for the next generation of research.
Library of Congress Subject Headings
Autonomous robots--Dynamics; Rotational motion; Inertial navigation; Accelerometers--Calibration
Mechanical Engineering (MS)
Department, Program, or Center
Mechanical Engineering (KGCOE)
Wayne W. Walter
Agamemnon L. Crassidis
Carr, Douglas A., "Development and Test of a Kinematical Method for Estimating Rotation Using the Acceleration of a Rigid Body" (2006). Thesis. Rochester Institute of Technology. Accessed from
RIT – Main Campus