Multi-agent interoperability and collaborative service mechanisms in heterogeneous environments and resource management

機構典藏 > College of Engineering > Graduate Institute & Department of Computer Science and Information Engineering > Thesis > Item 987654321/52354

Please use this identifier to cite or link to this item: https://tkuir.lib.tku.edu.tw/dspace/handle/987654321/52354

Title:	Multi-agent interoperability and collaborative service mechanisms in heterogeneous environments and resource management
Other Titles:	多代理人溝通及協同服務機制於異質環境與資源管理
Authors:	林益弘;Lin, Yi-hung
Contributors:	淡江大學資訊工程學系博士班葛煥昭;Keh, Huan-chao
Keywords:	多代理人;異質環境;資源管理;Multi-agent;Heterogeneous Environment;Resource Management
Date:	2010
Issue Date:	2010-09-23 17:34:16 (UTC+8)
Abstract:	近幾年來，在資訊科技急遽發展之下，資訊產品的功能已不再只是提供單一的功能，而是具備多重的功能，例如手持式電話不單單只是具備傳統式的電話功能，而是一個小型的智慧型電腦系統，有鑑於此一趨勢下，我們的研究是在這些系統中運用多代理人的特性自動達成溝通行為與資訊的傳遞、改進傳統的流程、減少管理成本以及提供更加靈活的資源運用。因此，在本論文中我們提出四種多代理行為，包括合作行為、溝通行為、協調行為及競爭行為，並且說明在多代理人中的溝通以及資源靈活運用策略ARAS(包括Amplification (用戶數量是大的，並且資源可能得到更多) 、Restriction (用戶數量是大的，並且資源不可能得到更多) 、Adjustment (數量是少量的用戶，並且資源是更多)以及Surveillance (用戶數量是少量，並且資源也是較少) )。此外，我們也使用了多個多代理人系統來說明多代理人行為與資源運用的結果，例如， Intelligent Parking System (IPS) ，IPS是一個整合了現有的RFID技術與多代理人行為，提供一個即時停車場狀態的訊息、 Resource Management System(RMS)，RMS導入了資源管理策略ARAS以及多代理行為，提供一個較靈活的資源管理系統和Mobile Learning Portal (MLP)，MLP是一個創新的互動式系統，包括學習、測試、警示、歷史資訊、社群及個人化。 In recent years, information technology (IT) has rapidly grown in popularity, with IT products beginning to provide more advanced functionality. For example, mobile phones now act not only as conventional phones but also as intelligent computer systems. In this vein, our research seeks to incorporate multi-agent characteristics into such products so that they can automatically execute tasks in the background, transmit messages, improve traditional processes, and reduce management costs, and thus provide more flexibility in the management of resources. Hence, we propose four types of multi-agent behavior—cooperation, communication, coordination, and competition—to realize a degree of interoperability among agents, as well as an amplification (large user volumes and a significant volume of available resources), restriction (large user volumes and constrained resources), adjustment (low user volumes and a significant volume of available resources), and surveillance (low user volumes and constrained resources), i.e., ARAS, approach to flexible resource management. In addition, we apply the proposed concepts in suggested implementations of various multi-agent behavior systems (case studies): an intelligent parking system (IPS), which integrates existing RFID technology and concepts of multi-agent behavior to provide parking lot information in real time; a resource management system (RMS), which implements the ARAS strategy and concepts of multi-agent behavior to provide a more flexible resource management system; and a mobile learning portal (MLP), which is an innovative, interactive system that includes a number of agents for varying purposes (i.e., learning, testing, alarm, history, community, and personalization).
Appears in Collections:	[Graduate Institute & Department of Computer Science and Information Engineering] Thesis

Files in This Item:

File	Size	Format
index.html	0Kb	HTML	335	View/Open

Loading...