Off-campus Michigan Tech users: To download campus access theses or dissertations, please use the following button to log in with your Michigan Tech ID and password: log in to proxy server
Non-Michigan Tech users: Please talk to your librarian about requesting this thesis or dissertation through interlibrary loan.
Date of Award
2024
Document Type
Campus Access Master's Thesis
Degree Name
Master of Science in Mechanical Engineering (MS)
Administrative Home Department
Department of Mechanical Engineering-Engineering Mechanics
Advisor 1
Gordon Parker
Committee Member 1
Hassan Masoud
Committee Member 2
Shangyan Zou
Abstract
Applying control strategies to Point Absorber Wave Energy Converters (PA-WEC) increases system efficiency and therefore system viability as a large scale renewable energy source. Many control strategies exist that rely on regular waves. Ocean waves are irregular with multiple frequency components which limit the success of strategies relying on regular waves. Additionally, linear models are often used directly or indirectly in control strategy design and are unable to exploit possible nonlinear response behaviors.
In this work, a process is developed for neural network, feedforward control of WECS exposed to regular or irregular waves that fully exploit nonlinear response characteristics. Training data is created from optimal control solutions for a variety of regular and irregular waves. Prior to final neural network training, a method is introduced for determining an appropriate time-series predictor set including the number and location of measurement backvalues.
Although both feedback and feedforward strategies were considered it was found that feedforward control, based solely on wave elevation, was preferred. Neural network controllers were trained separately for regular and irregular waves, yet it was found that the regular wave neural network performed well even when the WEC was exposed to irregular wave excitation. Interestingly, in many cases the neural network control outperformed the optimal solution do to its lack of constraints that were imposed when forming optimal control solutions.
Recommended Citation
Van Wieren, Madelyn G., "Neural Network Control of a Nonlinear Point Absorber Wave Energy Converter", Campus Access Master's Thesis, Michigan Technological University, 2024.