碩士 / 國立成功大學 / 造船工程學系研究所 / 85 / The world's ocean covers approximately 72 percent of the globe. There
are many projects about developing seas in advanced countries. It's vary
dangerous to work in underwater, so we develop many various automatic
machines to replace humans. For example, the Remoted Operated Vehicle (ROV).
The model of ROV is a complex math model, which owns six degree of
freedom with high nonlinerity and coupling. We try to design a controller
without exact system mode. Fuzzy control is a practicable way, it possesses
rules base similar to human's operation experience, using these to control
a plant.
The structure of the fuzzy controller is artificial neural networks. We
make use of the learning ability of artificial neural networks to adjust
proper membership functions. Generally, when we design a fuzzy controller
controller, we should use try-and-error to design the membership functions
of each linguistic variable. Artificial neural networks can solve this
problem. In order to simulate the motion of the ROV, first we should know
its dynamic mode and cable mode. Finally we observe controller's effect
by the motion simulation.
Identifer | oai:union.ndltd.org:TW/085NCKU0345018 |
Date | January 1996 |
Creators | Hsu, Li Ho, 許立和 |
Contributors | Huang Cheng Ching, 黃正清, --- |
Source Sets | National Digital Library of Theses and Dissertations in Taiwan |
Language | zh-TW |
Detected Language | English |
Type | 學位論文 ; thesis |
Format | 89 |
Page generated in 0.0223 seconds