碩士 / 國立成功大學 / 系統及船舶機電工程學系 / 103 / The main goal of the thesis is to find a feasible planning method of the track keeping for the AUV and the motion behaviors of the AUV in different environments through the Artificial Neural Network control system will be discussed. By controlling the AUV’s thrust, we can make the AUV’s motion be stable and reach its desired path. Based on the previous research of PMM (Planar Motion Mechanism) test for the AUV, we can provide the related hydrodynamic coefficients to the numerical computer program to solve the motion behaviors of the AUV, which can serve as the basis of the control system developed here. In the study, the LOS (Line of Sight) technique is applied to guide the heading of the AUV and the PD (Proportional- Derivative) controller incorporating with the Artificial Neural Network algorithm is adopted to allocate the proportional factors of the controller. Two kinds of the track keeping methods are adopted. The first is the traditional line of sight method and the second is applying the depth control based on the expected pitch angle. Both methods are also improved by using the step by step technique in order to obtain the more stable track keeping behaviors. Furthermore, the current effect is also included in the present study. From the present numerical simulation results, the neural network self-tuning PD controller is indeed more efficient on the AUV track keeping control than the traditional one. Besides, the second method with the expected depth control submitted here is also proved more stable than the traditional LOS method, especially for the 3D track keeping problem.
| Identifer | oai:union.ndltd.org:TW/103NCKU5345001 |
| Date | January 2014 |
| Creators | Chia-HsiangHung, 洪嘉祥 |
| Contributors | Ming-Chung Fang, Joe-Ming Yang, 方銘川, 楊澤民 |
| Source Sets | National Digital Library of Theses and Dissertations in Taiwan |
| Language | zh-TW |
| Detected Language | English |
| Type | 學位論文 ; thesis |
| Format | 102 |
Page generated in 0.0015 seconds