Detail Adaptive neural fuzzy logic control and its applications Bibliografi Author: Lee, E. Stanley (Advisor); Liu, Da-Ming Topik: COMPUTER SCIENCE Bahasa: (EN )    ISBN: 0-599-61318-1     Penerbit: KANSAS STATE UNIVERSITY     Tahun Terbit: 1999     Jenis: Theses - Dissertation Fulltext: 9958038.pdf (0.0B; 2 download) Abstract The neuro-fuzzy concept is a powerful modeling and control tool, which has been applied to many practical systems. The approach is especially useful for large complex and vague systems, which cannot be defined or represented reasonably as simple and unique. Thus, the approach is ideally suited for investigating the complex control problems of aircraft operations and other vaguely defined systems. The purpose of this research is to investigate these applications and to assess their capabilities and usefulness by actually simulating the system. When the aircraft is in the air, there are too many factors, which are generated by the atmosphere and the aircraft aerodynamics, to be controlled. These factors are generally called noises, which are vague, unknown and certainly not uniquely definable. The combined use of the learning ability of neural networks and the ability of representation of fuzzy systems can, at least, partially overcome this vagueness. The neuro-fuzzy network, which is known as the adaptive neural fizzy inference system (ANFIS) is applied to the to the control of aircraft operations, in particular the simulation of the autolanding trajectory tracking system of the aircraft. The goal is to design a system that can approximate the operation and then improve this approximation by neural network learning. The results prove that this is a very power approach, both for pilot training proposed simulation and for actual control. We also studied different learning techniques and discovered that the backpropagation learning, although slow, is the most reliable one due to its convergent property. The neuro-fuzzy system is also used to model two other vague systems, namely, thermal comfort and group technology. The former is a very vague system; in fact, even the term of thermal comfort is difficult to define. The latter is a very useful manufacturing concept but difficult to carry out. This research shows that both systems can be modeled successfully by the combined use of fuzzy modeling and neural network learning. Opini Anda Klik untuk menuliskan opini Anda tentang koleksi ini! Lihat Sejarah Pengadaan  Konversi Metadata   Kembali

Copyright © 2006, 2007 Unika Atma Jaya, all rights reserved