English  |  正體中文  |  简体中文  |  Items with full text/Total items : 26988/38789
Visitors : 2317477      Online Users : 27
RC Version 4.0 © Powered By DSPACE, MIT. Enhanced by NTU Library IR team.
Scope Adv. Search
LoginUploadHelpAboutAdminister

Please use this identifier to cite or link to this item: http://ntour.ntou.edu.tw:8080/ir/handle/987654321/33653

Title: 適應性智慧型系統於自走車導引控制之應用(III)
Authors: 莊季高
Contributors: 國立臺灣海洋大學:通訊與導航工程學系
Keywords: 適應性控制;小腦模型控制器;類神經網路;模糊系統;基因演算法;FPGA;定位系統;自動導航;路徑規劃;自走車
Adaptive Control;CMAC;Neural Network;Fuzzy System;Genetic Algorithm;GPS;FPGA;Localization System;Automatic Guidance;Path Planning;Autonomous Vehicle
Date: 2012-08
Issue Date: 2013-05-08T03:01:46Z
Publisher: 行政院國家科學委員會
Abstract: 摘要:本研究是結合適應控制及智慧型系統,選擇適合之類神經網路、基因演算法、模 糊系統於控制器之設計與應用,最主要的目的是要尋找具備適應環境變化能力之適應 性控制法則,建立結構簡單、學習快速的適當演算方式,以便在硬體實現時,可使得 成本降至最低。本研究應用一簡便型自走車,Dr Robot X80,結合超音波距離感測模 組、紅外線測距感測器、人體運動感測模組、影像模組、雷射測距感測器、GPS 等, 進行自動導航、尋跡、避撞、追車、超車、巡邏、巡航等適應性控制之應用。在導航 設計上以實用性能為設計之考量,如定位精確性、影像導航處理速度、通過狹宰空間 之能力等。在研究方法上加強系統定位之研究,並設計可進行快速有效之路徑規劃與 自動導航控制器。本計畫分三年完成,第一年(99/8-100/7)推導出 Adaptive Recurrent Neural Network (ARNN) 及 Adaptive Resource Allocating Network (ARAN) 學習法 則,應用於追跡控制上,並利用基因演算法探討最佳化能量控制。在硬體整合方面, 結合視覺與距離感測元件,簡化影像處理技術,以提高影像導航處理速度。第二年 (100/8-101/7) ( 執行中) 推導出適合於自走車之Adaptive CMAC 及 Adaptive GBF-CMAC 控制法則,並利用 Hopfield-Tank Neural Network 以轉折點概念探討具室 內動態避撞之最佳化路徑規劃方式。另外亦結合本實驗室已研發之路邊停車及倒車入 庫的自動控制功能(96-98 國科會計畫),應用於自走車室內清潔服務、室內公文傳遞 等的控制上。在硬體整合方面,結合視覺與距離感測元件與影像處理技術,應用嵌入 式系統並結合紅外線室內定位感測系統,提升定位精確性,實現於自走車室內導引控 制上。第三年(101/8-102/7)結合模糊系統於類神經網路及CMAC,推導出 Adaptive Type-1 與 Type-2 Fuzzy CMAC 及 Adaptive Linguistic Fuzzy Neural Network (ALFNN) 與 Adaptive Functional Fuzzy Neural Network (AFFNN) 學習法則。另外亦 結合第一年及第二年系統整合成果並加入GPS,應用於室外追車跟隨、超車、多車隊 形、動態避撞控制及通過狹宰空間上。在硬體整合方面,則以DSP/FPGA 實現適應 性智慧型車輛自動導引控制器之設計與製作。
abstract:This research integrates adaptive control and intelligent systems, such as neural networks, genetic algorithms, and fuzzy systems, to autonomous vehicle guidance and control. The main purposes of this project are to develop adaptive learning rule for the vehicle control system that can adopt environment disturbance and to construct simple and fast learning algorithm that can lower hardware cost. This research utilizes a wheel mobile robot called the Dr Robot X80 with ultrasonic sensor, inferred sensor, motion sensor, camera, laser ranger, and GPS to perform automatic guidance, trajectory tracking, obstacle avoidance, cut in, patrol, and cruise control. This is a three-year project. At the first year (2010.8-2011.7), we have derived learning rules for adaptive recurrent neural network and adaptive resource allocating network and apply it to chasing and following control. Genetic algorithm is used to find optimal energy control. We have integrated visual sensor, range sensor, image process, and embedded system to realize hardware control system. In the second year (2011.8-2012.7), we will derive adaptive CMAC and adaptive GBF-CMAC control rules for the autonomous vehicle. Hopfield-Tank neural network will be applied to dynamic obstacle avoidance and path planning. With parking control ability from previous work, the proposed autonomous vehicle will be able to perform floor cleaning and documentation delivering services. An inferred localization system will be used for indoor vehicle guidance and control. In the third year (2012.8-2013.7), we will derive proper learning rules of adaptive type-1 and type-2 fuzzy CMACs, adaptive linguistic fuzzy neural network, and adaptive functional fuzzy neural network for the vehicle control system. GPS will then be integrated in the control system for outdoor autonomous vehicle guidance and control tasks. Hardware controller will be implemented in a DSP/FPGA module. The feasibility of the proposed adaptive intelligent control system will be evaluated by the hardware test.
Relation: NSC101-2221-E019-038
URI: http://ntour.ntou.edu.tw/handle/987654321/33653
Appears in Collections:[通訊與導航工程學系] 研究計畫

Files in This Item:

There are no files associated with this item.



All items in NTOUR are protected by copyright, with all rights reserved.

 


著作權政策宣告: 本網站之內容為國立臺灣海洋大學所收錄之機構典藏,無償提供學術研究與公眾教育等公益性使用,請合理使用本網站之內容,以尊重著作權人之權益。
網站維護: 海大圖資處 圖書系統組
DSpace Software Copyright © 2002-2004  MIT &  Hewlett-Packard  /   Enhanced by   NTU Library IR team Copyright ©   - Feedback