Title page for etd-0722119-171110

URN	etd-0722119-171110	Statistics	This thesis had been viewed 197 times. Download 3 times.
Author	Chih-Hsiang Lin
Author's Email Address	satan.angel.god@gmail.com
Department	Institute Of Mechanical Engineering
Year	2018	Semester	2
Degree	Master	Type of Document	Master's Thesis
Language	zh-TW.Big5 Chinese	Page Count	79
Title	The Artificial Neural Emulator Design and Analysis for DC Motor Systems
Keyword	Neural Network Control DC Motor Diagonal Recurrent Neural Network Neural Network Emulator Neural Network Emulator Diagonal Recurrent Neural Network DC Motor Neural Network Control
Abstract	Generally, multi-layer neural network controllers are used for DC motor speed control. Most of them use indirect neural network adaptive control. To use this control architecture, an offline or online neural network emulator must be designed. The emulator must be pre-trained offline. Tested and expected to increase the rate of convergence. 　　In this study, a three-layer neural network and a diagonal recurrent neural network are used to design a DC motor-like neural network emulator. It is expected to have a fast convergence rate and properly determine the system sensitivity. The simulation results show that the proposed method is applied to the DC motor-like neural network emulator, which can obtain fast convergent speed and establish a stable neural network control system.
Advisor Committee	Ming-hui Chu - advisor Cheng-kuang Huang - co-chair Hsiao-chung Liu - co-chair
Files	50602007.pdf indicate access worldwide
Date of Defense	2018-07-10	Date of Submission	2019-07-23

Back to Results