Fiber Optic Sensors for Liquid Identifications

<p>"The fiber optic Fabry-Perot interferometer (FPI) is a widely utilized sensing configuration, offering numerous advantages over conventional electronic sensors, including high accuracy, distributed sensing capabilities, immunity to electromagnetic interference, and compact size. In this study, we propose a remarkably simple fiber optic-tip sensor system combined with machine learning techniques for the identification of pure and volatile organic liquids (VOLs).</p> <p>A liquid droplet forms an extrinsic FPI (EFPI), with its effective reflectance being a function of the droplet's length. As the droplet evaporates, its length decreases. We conducted immersion tests using optical fiber tip sensors and monitored the time-transient responses of the evaporating droplets. Inspired by the evaporation dynamics of liquids, we employed machine learning techniques to efficiently extract valuable information from the evaporation time-transient signals of liquid pendant droplets. The time-transient signal was converted into image data using a continuous wavelet transform, and convolutional neural network (CNN) models were then applied to predict the liquid being tested based on the image data. Consequently, we developed a sensing system utilizing advanced data-driven techniques, such as machine learning, for liquid identification.</p> <p>This innovative and intelligent sensor system has the potential to serve as a foundation for a new generation of powerful sensor networks"--Abstract, p. iv</p>

Name thesis:degree_name

Ph. D. in Electrical Engineering

Grantor

Missouri University of Science and Technology

Author dc:creator

• Naku, Wassana

• Electrical and Computer Engineering
• Engineering

Repository record dc:identifier

https://scholarsmine.mst.edu/doctoral_dissertations/3257

OAI identifier oai:identifier

oai:scholarsmine.mst.edu:doctoral_dissertations-4262