Sensor Characterization and Signal Fusion for InstantEye

Author

Wyman T. Smith, University of New Hampshire, DurhamFollow

Date of Award

Spring 2017

Project Type

Senior Honors Thesis

College or School

CEPS

Department

Physics

Program or Major

Engineering Physics

Degree Name

Bachelor of Science

First Advisor

Marc Lessard

Abstract

The practicality and effectiveness of using a TerraRanger Duo—a parallel sonar and infrared time-of-flight distance sensor—payload for obstacle detection is investigated for use with Physical Science Inc.’s InstantEye drone. A Python program was developed to interface with the serial data output before comparing the sensor’s empirical performance against its data sheet. The two signals from the distinct sensor modules, each with their characterized strengths and weaknesses, were then fused with a Kalman filter. This was further refined by imposing conditional weighting based on the known sensor characteristics. The filter output, with conditional corrections, was able to accurately track a single object’s position and velocity within a maximum range of 14 meters.

Recommended Citation

Smith, Wyman T., "Sensor Characterization and Signal Fusion for InstantEye" (2017). Honors Theses and Capstones. 362.
https://scholars.unh.edu/honors/362