Portland State University. Department of Electrical and Computer Engineering
Date of Publication
Master of Science (M.S.) in Electrical and Computer Engineering
Electrical and Computer Engineering
Ocean bottom -- Classification -- Remote sensing, Underwater acoustics -- Measurement, Underwater acoustics -- Mathematical models
1 online resource (viii, 93 pages)
As of the year 2019, only about five percent of the seafloor has been topologically mapped and classified for type (e.g., sand, silt, gravel). To rapidly survey the seabed from surface ships or underwater vehicles, acoustic remote sensing methods are needed. In this thesis, acoustic measurements and inverse modeling is investigated as a way to classify seabed type based on estimating parameters such as density, sound speed, and interface roughness. The method uses normal incident acoustic measurements that can be made using either a single beam echo sounder or the normal incident beams from a side scan sonar. The inverse method consists of a forward model to simulate the signal and a directed search over parameter space based on an evolutionary algorithm. To direct the search, the similarity between modeled and data envelopes are quantified using a Huber Loss objective function. The large search space requires rapid forward model calculations so a ray-based model was implemented. To determine the applicability of the ray-based model at low frequencies, a validation study is presented to compare the approximate ray method against exact solutions. The full inversion problem was also considered for sediment types of sand, silt and gravel at various signal to noise ratios and those results are included.
In Copyright. URI: http://rightsstatements.org/vocab/InC/1.0/ This Item is protected by copyright and/or related rights. You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s).
Frantz, Megan, "Classifying Seabed Parameters from Normal Incidence Reflections: Model Comparison and Inversion Technique" (2020). Dissertations and Theses. Paper 5396.